- **Codex no longer re-registers the Claude SessionStart hook.** v6.1.0 removed the Codex hook config and its manifest `hooks` pointer, meaning to stop Codex from installing a SessionStart hook — but with no `hooks` field, Codex fell back to auto-discovering `hooks/hooks.json`, the Claude Code SessionStart hook that the marketplace ships from the repo root, and re-registered it along with its install-time trust prompt. The Codex manifest now declares an explicit empty hooks object (`hooks: {}`), which Codex reads as "no hooks" instead of reaching the auto-discovery fallback. An absent field, `[]`, and an empty inline list all collapse back to the fallback, so the value has to be exactly `{}`.
- **Removed orphaned Codex session-start dead code.** `hooks/session-start-codex` had no caller once the Codex hook config was deleted, so it and its redundant test cases are gone. The worked shell-hook example in `docs/porting-to-a-new-harness.md` moves from Codex — now native skill discovery with no session-start hook — to Cursor, a live shell-hook harness, and the stale `hooks-codex.json` pointer in `docs/windows/polyglot-hooks.md` is corrected. The Codex plugin category is also fixed to "Developer Tools".
### Packaging
- **New `package-codex-plugin.sh` for building the Codex portal package.** A maintainer script produces a deterministic Codex "portal" archive — `.zip` by default, `tar.gz` on request — that normalizes entry timestamps, preserves executable modes, verifies every packaged skill ships its OpenAI metadata, includes the app and composer icons, and refuses to run against a dirty worktree. The packaged manifest keeps the source `hooks: {}` object so a portal-installed plugin avoids the same SessionStart auto-discovery, and the script can rebuild a byte-identical archive from a saved metadata source. Covered by a new test suite.
The `using-superpowers` bootstrap is injected into every session, so its size is paid for constantly. This release trims it and the per-harness references it points to, without dropping behavior-shaping content.
- **Compressed the `using-superpowers` bootstrap.** Replaced the graphviz skill-flow diagram with the prose it encoded, folded the standalone Instruction-Priority section into User Instructions, dropped the per-platform "How to Access Skills" walkthrough, and trimmed the Platform Adaptation pointer to the harnesses that still ship a reference file. The full Red Flags rationalization table and the user-instruction precedence rules are unchanged.
- **Pruned the per-harness tool-mapping references.** The verbose action-to-tool tables restated guidance modern agents already follow. Each reference file is trimmed to the harness-specific notes that still carry weight — subagent dispatch, task tracking, instructions-file paths — and `claude-code-tools.md` and `copilot-tools.md`, which had nothing harness-specific left, are deleted.
### Codex
- **Codex can install from the marketplace.** Codex marketplace sources expect a `.agents/plugins/marketplace.json` at the marketplace root; the repo only shipped the Claude marketplace file, so Codex could name the marketplace but found no installable plugin entries. A repo-local Codex marketplace manifest now points at the same repository root, so the plugin is installable from Codex.
- **Codex no longer ships a SessionStart hook.** Codex reliably triggers skills on its own, and the bootstrap hook made the UX worse rather than better. The Codex hook config (`hooks-codex.json`) and its manifest registration are removed.
### Harness Support
- **Gemini CLI support removed.** Google EOLed the Gemini CLI on 2026-06-18; the extension can no longer be installed or updated. Gemini is gone from the install docs, the subagent-capable platform lists, and the eval-harness description, and its tool-mapping reference is deleted.
- **SDD scratch files moved out of `.git/`.** Claude Code treats `.git/` as a protected path and denies agent writes there, so an implementer subagent writing its report into `.git/sdd/` got blocked mid-run. Task briefs, implementer reports, review diffs, and the progress ledger now live in a self-ignoring `.superpowers/sdd/` directory in the working tree — kept out of `git status` and out of commits, and resolved per worktree by a shared `sdd-workspace` helper. One caveat: because the workspace is git-ignored working-tree scratch, `git clean -fdx` will delete the progress ledger; recover from `git log` if that happens. (#1780)
- **We no longer ship the `evals` submodule.** It broke plugin installs for some users, so the eval harness now lives in its own repo, separate from the published plugin. (#1778, #1774)
- **Version display in the brainstorm companion** — packaged Codex plugins ship without a root `package.json`, so the visual companion reported its version as "unknown". `readSuperpowersVersion()` now falls back to `.codex-plugin/plugin.json` when `package.json` is absent.
- **Cleaner Codex plugin sync** — the sync-to-codex script now excludes `.gitmodules` and `.pre-commit-config.yaml`, keeping repo metadata out of the packaged Codex plugin.
Superpowers 6.0 is a big release. The headline is a rewrite of how `subagent-driven-development` reviews each task — cheaper, stricter, and harder to game.
While these numbers won't hold on every harness and for every workload, in our evals, Claude Code and Codex produce similar high-quality results roughly twice as fast and while spending almost 50% fewer tokens.
It also adds three new harnesses (Kimi Code, Pi, and Antigravity), gives the brainstorming visual companion a better security model, and rewrites a number of skills' tool calls to be significantly more vendor-neutral.
### Visible Changes
- **The two per-task reviewer prompts became one.** `spec-reviewer-prompt.md` and `code-quality-reviewer-prompt.md` are gone, replaced by a single `task-reviewer-prompt.md`. If you dispatch the old files directly, switch to the new one.
- **The legacy global worktree directory is gone.** `using-git-worktrees` and `finishing-a-development-branch` no longer use `~/.config/superpowers/worktrees/`. Worktrees now land in the project — an existing `.worktrees/` or `worktrees/` if you have one, otherwise a fresh `.worktrees/` — unless you say otherwise.
### New Harness Support
Superpowers now runs on three more harnesses. Each ships its own bootstrap, a tool-mapping reference, and tests, and each gets its own install section in the README.
- **Kimi Code** — a plugin manifest, install docs, and manifest tests; install from Kimi's marketplace or straight from the repo. (initial manifest by @qer)
- **Pi** — a session-start extension that registers the skills and injects the `using-superpowers` bootstrap. Pi has native skills, so it needs no compatibility shim.
- **Antigravity (`agy`)** — installs the plugin directly and bootstraps from the first message; verified end-to-end against the standard "make a react todo list" acceptance test.
### Subagent-Driven Development
A long run of cost-and-quality experiments on real projects reshaped how the controller reviews each task. The old flow ran two reviewers per task and leaned on the controller's judgment for model choice and severity, and both turned out to be expensive and easy to game. The new flow runs one reviewer per task, hands work off as files instead of pasted text, and takes several judgment calls away from the controller.
- **One reviewer per task, two verdicts.** A single `task-reviewer-prompt.md` reads the task's diff once and returns both a spec-compliance verdict and a quality verdict, so one fix pass clears both. A new "can't verify from the diff" verdict flags requirements that live in untouched code, for the controller to check itself. (#1538, #1543)
- **One broad review at the end.** The run finishes with a single whole-branch review on the most capable model, instead of re-reviewing everything task by task.
- **Plans get a pre-flight read.** Before the first task, the controller checks the plan for internal conflicts — and for anything the plan asks for that a reviewer would flag as a defect — and raises it all at once, rather than stumbling into it mid-run.
- **Diffs and task text move as files.** A pasted diff parks itself permanently in the most expensive context, and a reviewer without one rebuilds it by hand — the single biggest reviewer cost. Two new scripts, `task-brief` and `review-package`, write the task text and the review diff to files for the subagent to read.
- **Every dispatch states its model.** Left to choose, controllers stopped naming a model at all — and an unnamed model quietly inherits the session's most expensive one, so one run put all 26 of its reviewers on the top tier. The templates now require a model, with guidance that reaches for cheaper tiers when the work allows.
- **The controller can't tell a reviewer what to ignore.** Real runs caught controllers coaching reviewers to skip a finding or call it "Minor at most," and the flaw shipped. Suppressing findings and pre-rating severity are now banned outright, and a defect the plan itself mandates gets reported for you to decide on rather than waved through.
- **Reviewers are read-only and skeptical of rationales.** Review no longer touches the working tree or branch — a reviewer running `git checkout` had been orphaning later commits — and an implementer's "I left this unabstracted on purpose" no longer talks a reviewer out of a real finding.
- **Stronger evidence and reporting.** Reviewers back each answer with a file and line, the implementer's report moves to a file and carries red/green evidence when TDD applies, and a progress ledger lets a controller that loses its context resume instead of redoing finished work. (#994)
### Writing Plans
Plans now carry the structure the controller and reviewers used to re-derive on every dispatch.
- **A Global Constraints block** lists the rules that bind every task — version floors, dependency limits, naming and copy, exact values — copied in verbatim, so they actually reach the implementers and reviewers downstream.
- **A per-task Interfaces block** names exactly what each task consumes and produces, so an implementer who sees only its own task still knows its neighbors' contracts.
- **Right-sizing guidance** keeps a task at the size that earns its own test cycle and a reviewer's pass, folding setup, config, and docs into the task that needs them. In testing, a plan written this way needed one round of fixes where the control needed two to four — and the control shipped a real bug.
### Brainstorming Visual Companion
The visual companion is a small web server the agent opens alongside the conversation. It had no authentication at all, so on a shared or remote machine anyone who could reach the port could read your brainstorm — or inject events the agent treats as your input. This release gives it a real security model and makes it survive restarts and dropped connections.
- **A per-session key now guards everything.** The agent's URL carries a one-time key, the browser tucks it into a tab-scoped cookie, and every request and WebSocket connection has to present it. This closes the door to stray local tabs and routable remote hosts alike, including the DNS-rebinding case an origin allowlist can't catch. (Closes #1014)
- **The file server stays in its sandbox.** It refuses symlinks, dotfiles, and any path that climbs out of the content directory, ignores macOS resource-fork files, and sends the usual no-store and deny-framing headers. Files that hold the session key are written owner-only.
- **The companion is offered only when it helps.** The skill raises it the first time a question would read better shown than told, as its own message, and lets a decline stand. Accepting opens your browser to the first screen. (Closes #755)
- **It survives restarts and flaky connections.** Given a project directory, the server keeps the same port and key across restarts, so an open tab simply reconnects. The page reconnects on its own, shows a live status pill, and raises a "paused" overlay while the server is down.
- **Longer idle life, safer shutdown.** The idle timeout went from 30 minutes to 4 hours, and `stop-server.sh` now confirms it owns the right process before signaling, so it never kills an unrelated `node` after a reboot. (#1703)
- **Windows launch hardening** — consolidated shell detection, and Windows now relies on the idle timeout for shutdown, since Node can't track POSIX process ownership across MSYS2.
### Existing Harness Updates
- **Codex** now bootstraps through its own SessionStart hook rather than shared wiring, and the Codex App gained an install section and fuller tool docs (web search, `AGENTS.md`, personal skills). (#1540)
- **OpenCode** got an action-based tool mapping across its plugin, install doc, and README, plus a bootstrap-caching test.
- **Cursor**'s manifest dropped its `agents` and `commands` entries, since those directories no longer exist.
### One Set of Skills, Every Harness
The skills used to speak Claude Code's dialect — "use the Task tool," "put it in CLAUDE.md." This release rewrites that vocabulary in terms of what you're actually doing ("dispatch a subagent," "your instructions file") and adds a per-harness reference that maps each action to the right tool, checked against each runtime. Prose that named "Claude" now says "your agent."
- **A tool reference per harness** at `skills/using-superpowers/references/`, covering Claude Code, Codex, Copilot, Gemini, Pi, and Antigravity.
- **`finishing-a-development-branch` went forge-neutral** — it no longer hardcodes `gh pr create`, so agents push with whatever forge tooling they have. (#1609)
- **One rename:** "Claude Search Optimization" is now "Skill Discovery Optimization," since the technique isn't Claude-specific.
### Writing Skills
Two additions for skill authors.
- **Match the Form to the Failure** — a short table for picking the right kind of guidance. A flat "don't do X" works for discipline slips but backfires when the problem is the *shape* of an output, where a worked example does better. The table, and a tighter scope on the existing rationalization section, steer authors to the form that actually helps.
- **Micro-Test Wording** — a cheap way to check a phrasing before committing to it: sample it a handful of times against a no-guidance control and read every result by hand, treating run-to-run variance as a warning sign.
### Testing
Skill-behavior testing moved out of `tests/` into a new `evals/` submodule built on "drill," which runs real Claude Code, Codex, and Gemini sessions and judges them with an LLM. Several in-tree bash suites retired once a stricter drill scenario covered them; the few with no equivalent stayed. From here on, `tests/` holds plugin-code tests and `evals/` holds skill-behavior tests, and `docs/testing.md` explains the split. New backends reach Antigravity, Pi, and more models, and new shell-lint and pre-commit checks guard the harness. (#1541)
### Bug Fixes
- **systematic-debugging no longer forces every session into extended thinking.** One bullet held the exact keyword Claude Code scans for, quietly tripping the switch on every session that loaded the skill. A hyphen breaks the keyword; the text still reads. (#1283, by @Nick Galatis)
- **The Windows SessionStart hook stopped printing a write error every session** — each `printf` now routes through `cat` to absorb the broken pipe, and the output is otherwise unchanged. (#1612, reported by @silvertakana)
- **Windows foreground mode** tracks the right process and clears its owner PID on MSYS2. (by @nestorluiscamachopaz)
- **The `using-superpowers` bootstrap** no longer lists "debugging" as a skill that doesn't exist. (reported by @mhat)
- **The TDD skill** links the testing anti-patterns reference. (#1532, #1529; link fix #1474 by @Stable Genius)
- **`using-git-worktrees`** fixes its step numbering and drops stale Cursor references. (#1522, and by @fuleinist)
- **The Codex review skill** swaps a private in-joke for plain guidance. (#1531)
### Documentation & Contributor Guidelines
- **A guide to porting Superpowers to a new harness** (`docs/porting-to-a-new-harness.md`) lays out the three pieces every integration needs and the one rule that makes or breaks it: load the bootstrap at session start.
- **Every PR and issue now discloses how it was made** — model, harness, version, and installed plugins, or a note that it was written by hand. We weigh a contribution differently depending on what produced it. PRs also target `dev`, not `main`. The PR template, all three issue templates, and a new platform-support template carry this.
### Contributors
Thanks to @mattvanhorn, @nawfal, @Nick Galatis, @silvertakana, @nestorluiscamachopaz, @qer, @mhat, @Stable Genius, @fuleinist, @dev_Hakaze, @robotsnh, Rahul, and @arittr.
- **Legacy slash commands removed** — `/brainstorm`, `/execute-plan`, and `/write-plan` are gone. They were deprecated stubs that did nothing but tell the user to invoke the corresponding skill. Invoke `superpowers:brainstorming`, `superpowers:executing-plans`, and `superpowers:writing-plans` directly instead. (#1188)
- **`superpowers:code-reviewer` named agent removed** — the agent was the plugin's only named agent and was used by exactly two skills, while every other reviewer/implementer subagent in the repo dispatches `general-purpose` with a prompt template alongside its skill. The agent's persona and checklist have been merged into `skills/requesting-code-review/code-reviewer.md` as a self-contained Task-dispatch template. Anyone dispatching `Task (superpowers:code-reviewer)` should switch to `Task (general-purpose)` with the prompt template instead. (PR #1299)
- **Integration sections removed from skills** — these were a legacy of the time before agents had native skills systems and didn't help with steering.
### Worktree Skills Rewrite
`using-git-worktrees` and `finishing-a-development-branch` now detect when the agent is already running inside an isolated worktree and prefer the harness's native worktree controls before falling back to `git worktree`. Behavior was TDD-validated and cross-platform-checked across five harnesses. (PRI-974, PR #1121)
- **Environment detection** — both skills check `GIT_DIR != GIT_COMMON` before doing anything; if already in a linked worktree, creation is skipped entirely. A submodule guard prevents false detection.
- **Consent before creating worktrees** — `using-git-worktrees` no longer creates worktrees implicitly; the skill asks the user first. Fixes #991 (subagent-driven-development was auto-creating worktrees without consent).
- **Native tool preference (Step 1a)** — when the harness exposes its own worktree tool (e.g. Codex), the skill defers to it. The user's stated preference is respected when expressed.
- **Provenance-based cleanup** — `finishing-a-development-branch` only cleans up worktrees inside `.worktrees/` (created by superpowers); anything outside is left alone. Fixes #940 (Option 2 was incorrectly cleaning up worktrees), #999 (merge-then-remove ordering), and #238 (`cd` to repo root before `git worktree remove`).
- **Detached HEAD handling** — the finishing menu collapses to two options when there is no branch to merge from.
- **Hardcoded `/Users/jesse` paths** in skill examples replaced with generic placeholders. (#858, PR #1122)
### Contributor Guidelines for AI Agents
Two new sections at the top of `CLAUDE.md` (symlinked to `AGENTS.md`) speak directly to AI agents. An audit of the last 100 closed PRs against this repo showed a 94% rejection rate driven by AI-generated slop: agents that didn't read the PR template, opened duplicates, fabricated problem descriptions, or pushed fork- or domain-specific changes upstream.
- **Pre-submission checklist** — read the PR template, search for existing PRs, verify a real problem exists, confirm the change belongs in core, and show the human partner the complete diff before submitting.
- **What we will not accept** — third-party dependencies, "compliance" rewrites of skill content, project-specific configuration, bulk PRs, speculative fixes, domain-specific skills, fork-specific changes, fabricated content, and bundled unrelated changes.
- **New harness PRs require a session transcript** — most past new-harness integrations copied skill files or wrapped with `npx skills` instead of loading the `using-superpowers` bootstrap at session start. The acceptance test ("Let's make a react todo list" must auto-trigger `brainstorming` in a clean session) and a complete transcript are now required.
### Codex Plugin Mirror Tooling
New `sync-to-codex-plugin` script mirrors superpowers into the OpenAI Codex plugin marketplace as `prime-radiant-inc/openai-codex-plugins`. Path/user-agnostic so any team member can run it. (PR #1165)
- Clones the fork fresh into a temp directory per run, regenerates overlays inline, and opens a PR; auto-detects upstream from the script's own location and preflights `rsync`/`git`/`gh auth`/`python3`.
-`--bootstrap` flag for first-time setup; `EXCLUDES` patterns anchored to source root; `assets/` excluded.
- Mirrors `CODE_OF_CONDUCT.md`; drops the `agents/openai.yaml` overlay.
- Seeds `interface.defaultPrompt` in the mirrored `plugin.json`. (PR #1180 by @arittr)
- Codex plugin files are committed to the source repo so the sync script uses canonical versions; Codex marketplace metadata is preserved.
### OpenCode
- **Bootstrap content cached at module level** — `getBootstrapContent()` was calling `fs.existsSync` + `fs.readFileSync` + frontmatter regex on every agent step (the `experimental.chat.messages.transform` hook fires on every step in OpenCode's agent loop). Now read once, cached for the session lifetime, with a null sentinel for the missing-file case. 15 regression tests cover cache behavior, fs call counts, the injection guard, the missing-file sentinel, and cache reset. (Fixes #1202)
- **Integration tests modernized**.
- **Install caveats clarified** in the README.
### Code Review Consolidation
`requesting-code-review` is now self-contained: the persona, checklist, and dispatch template live in `skills/requesting-code-review/code-reviewer.md` and the skill dispatches `Task (general-purpose)` directly. (PR #1299)
- **Single source of truth** — the persona/checklist that previously lived in both `agents/code-reviewer.md` and the skill's placeholder template (and drifted independently) is now one file.
- **`subagent-driven-development` follows suit** — its `code-quality-reviewer-prompt.md` now dispatches `Task (general-purpose)` instead of the named agent.
- **Behavioral test added** — `tests/claude-code/test-requesting-code-review.sh` plants real bugs (SQL injection, plaintext password handling, credential logging) into a tiny project and asserts the dispatched reviewer flags every planted issue at Critical/Important severity and refuses to approve the diff.
> Note: `tests/claude-code/test-requesting-code-review.sh` and `tests/claude-code/test-document-review-system.sh` (mentioned later in this document) were lifted into drill scenarios on 2026-05-06 and removed from `tests/`. See `evals/scenarios/code-review-catches-planted-bugs.yaml` and `evals/scenarios/spec-reviewer-catches-planted-flaws.yaml`. The references above and below are preserved as dated artifacts of the work this section describes.
- **Codex and Copilot workaround docs trimmed** — the "Named agent dispatch" sections in `references/codex-tools.md` and `references/copilot-tools.md` documented how to flatten a named agent into a generic dispatch. With no named agents shipping, the workaround is unnecessary; both sections were dropped.
### Subagent-Driven Development
- **No more pause every 3 tasks** — the "review after each batch (3 tasks)" cadence in `requesting-code-review` (originally for `executing-plans`) was leaking into `subagent-driven-development`. Replaced with "each task or at natural checkpoints" plus an explicit continuous-execution directive.
- **SDD integration test now runs its assertions** — three independent bugs caused the test to silently bail before printing any verification results: an unresolved `..` segment in the working-dir path, a `set -euo pipefail` interaction with `find | sort | head -1` (SIGPIPE on the producer killed the script), and a missing `--plugin-dir` on the `claude -p` invocation that caused the test to load the installed plugin instead of the working tree. All three fixed; six verification tests now actually run against a real end-to-end SDD run.
### Cursor
- **Windows SessionStart hook** routed through `run-hook.cmd` instead of invoking the extensionless `session-start` script directly. Fixes Windows opening the file in an editor instead of running it. Also removed an accidental UTF-8 BOM from `hooks-cursor.json`.
### Gemini CLI
- **Subagent dispatch mapping** — Gemini's `Task` dispatch now maps to `@agent-name` / `@generalist`, with parallel subagent dispatch documented for independent tasks.
### Skills
- **Terminology cleanups** across skill content.
### Documentation & Install
- **Factory Droid installation instructions** added to README.
- **Quickstart install links** in README. (PR #1293 by @arittr)
- **Codex plugin install guidance** updated. (PR #1288 by @arittr)
- **Codex `wait` mapping corrected** to `wait_agent` in the tools reference.
- **Install order reorganized**; Codex install instructions cleaned up.
- **Removed vestigial `CHANGELOG.md`** in favor of `RELEASE-NOTES.md` as the single source. (PR #1163 by @shaanmajid)
- **Discord invite link** fixed; release announcements link and a detailed Discord description added to the Community section.
- **SessionStart context injection** — Copilot CLI v1.0.11 added support for `additionalContext` in sessionStart hook output. The session-start hook now detects the `COPILOT_CLI` environment variable and emits the SDK-standard `{ "additionalContext": "..." }` format, giving Copilot CLI users the full superpowers bootstrap at session start. (Original fix by @culinablaz in PR #910)
- **Tool mapping** — added `references/copilot-tools.md` with the full Claude Code to Copilot CLI tool equivalence table
- **Skill and README updates** — added Copilot CLI to the `using-superpowers` skill's platform instructions and README installation section
- **Skills path consistency** — the bootstrap text no longer advertises a misleading `configDir/skills/superpowers/` path that didn't match the runtime path. The agent should use the native `skill` tool, not navigate to files by path. Tests now use consistent paths derived from a single source of truth. (#847, #916)
- **Bootstrap as user message** — moved bootstrap injection from `experimental.chat.system.transform` to `experimental.chat.messages.transform`, prepending to the first user message instead of adding a system message. Avoids token bloat from system messages repeated every turn (#750) and fixes compatibility with Qwen and other models that break on multiple system messages (#894).
The subagent review loop (dispatching a fresh agent to review plans/specs) doubled execution time (~25 min overhead) without measurably improving plan quality. Regression testing across 5 versions with 5 trials each showed identical quality scores regardless of whether the review loop ran.
- **Session directory restructured** — the brainstorm server session directory now contains two peer subdirectories: `content/` (HTML files served to the browser) and `state/` (events, server-info, pid, log). Previously, server state and user interaction data were stored alongside served content, making them accessible over HTTP. The `screen_dir` and `state_dir` paths are both included in the server-started JSON. (Reported by 吉田仁)
- **Owner-PID lifecycle fixes** — the brainstorm server's owner-PID monitoring had two bugs causing false shutdowns within 60 seconds: (1) EPERM from cross-user PIDs (Tailscale SSH, etc.) was treated as "process dead", and (2) on WSL the grandparent PID resolves to a short-lived subprocess that exits before the first lifecycle check. Fixed by treating EPERM as "alive" and validating the owner PID at startup — if it's already dead, monitoring is disabled and the server relies on the 30-minute idle timeout. This also removes the Windows/MSYS2-specific carve-out from `start-server.sh` since the server now handles it generically. (#879)
- **writing-skills** — corrected false claim that SKILL.md frontmatter supports "only two fields"; now says "two required fields" and links to the agentskills.io specification for all supported fields (PR #882 by @arittr)
### Codex App Compatibility
- **codex-tools** — added named agent dispatch mapping documenting how to translate Claude Code's named agent types to Codex's `spawn_agent` with worker roles (PR #647 by @arittr)
- **codex-tools** — added environment detection and Codex App finishing sections for worktree-aware skills (by @arittr)
- **Brainstorm server ESM fix** — renamed `server.js` → `server.cjs` so the brainstorming server starts correctly on Node.js 22+ where the root `package.json``"type": "module"` caused `require()` to fail. (PR #784 by @sarbojitrana, fixes #774, #780, #783)
- **Brainstorm owner-PID on Windows** — skip PID lifecycle monitoring on Windows/MSYS2 where the PID namespace is invisible to Node.js, preventing the server from self-terminating after 60 seconds. (#770, docs from PR #768 by @lucasyhzlu-debug)
- **stop-server.sh reliability** — verify the server process actually died before reporting success. SIGTERM + 2s wait + SIGKILL fallback. (#723)
### Changed
- **Execution handoff** — restore user choice between subagent-driven and inline execution after plan writing. Subagent-driven is recommended but no longer mandatory.
- **Single whole-plan review** — plan reviewer now reviews the complete plan in one pass instead of chunk-by-chunk. Removed all chunk-related concepts (`## Chunk N:` headings, 1000-line chunk limits, per-chunk dispatch).
- **Raised the bar for blocking issues** — both spec and plan reviewer prompts now include a "Calibration" section: only flag issues that would cause real problems during implementation. Minor wording, stylistic preferences, and formatting quibbles should not block approval.
- **Reduced max review iterations** — from 5 to 3 for both spec and plan review loops. If the reviewer is calibrated correctly, 3 rounds is plenty.
- **Streamlined reviewer checklists** — spec reviewer trimmed from 7 categories to 5; plan reviewer from 7 to 4. Removed formatting-focused checks (task syntax, chunk size) in favor of substance (buildability, spec alignment).
- **One-line plugin install** — OpenCode plugin now auto-registers the skills directory via a `config` hook. No symlinks or `skills.paths` config needed. Install is just adding one line to `opencode.json`. (PR #753)
- **Added `package.json`** so OpenCode can install superpowers as an npm package from git.
### Bug Fixes
- **Verify server actually stopped** — `stop-server.sh` now confirms the process is dead before reporting success. SIGTERM + 2s wait + SIGKILL fallback. Reports failure if the process survives. (PR #751)
- **Cursor hooks** — added `hooks/hooks-cursor.json` with Cursor's camelCase format (`sessionStart`, `version: 1`) and updated `.cursor-plugin/plugin.json` to reference it. Fixed platform detection in `session-start` to check `CURSOR_PLUGIN_ROOT` first (Cursor may also set `CLAUDE_PLUGIN_ROOT`). (Based on PR #709)
- **Stop firing SessionStart hook on `--resume`** — the startup hook was re-injecting context on resumed sessions, which already have the context in their conversation history. The hook now fires only on `startup`, `clear`, and `compact`.
- **Bash 5.3+ hook hang** — replaced heredoc (`cat <<EOF`) with `printf` in `hooks/session-start`. Fixes indefinite hang on macOS with Homebrew bash 5.3+ caused by a bash regression with large variable expansion in heredocs. (#572, #571)
- **POSIX-safe hook script** — replaced `${BASH_SOURCE[0]:-$0}` with `$0` in `hooks/session-start`. Fixes "Bad substitution" error on Ubuntu/Debian where `/bin/sh` is dash. (#553)
- **Portable shebangs** — replaced `#!/bin/bash` with `#!/usr/bin/env bash` in all shell scripts. Fixes execution on NixOS, FreeBSD, and macOS with Homebrew bash where `/bin/bash` is outdated or missing. (#700)
- **Brainstorm server on Windows** — auto-detect Windows/Git Bash (`OSTYPE=msys*`, `MSYSTEM`) and switch to foreground mode, fixing silent server failure caused by `nohup`/`disown` process reaping. (#737)
- **Codex docs fix** — replaced deprecated `collab` flag with `multi_agent` in Codex documentation. (PR #749)
- Full test suite: HTTP serving, WebSocket protocol, file watching, and integration tests
### Brainstorm Server Reliability
- **Auto-exit after 30 minutes idle** — server shuts down when no clients are connected, preventing orphaned processes
- **Owner process tracking** — server monitors the parent harness PID and exits when the owning session dies
- **Liveness check** — skill verifies server is responsive before reusing an existing instance
- **Encoding fix** — proper `<meta charset="utf-8">` on served HTML pages
### Subagent Context Isolation
- All delegation skills (brainstorming, dispatching-parallel-agents, requesting-code-review, subagent-driven-development, writing-plans) now include context isolation principle
- Subagents receive only the context they need, preventing context window pollution
- Per-platform launch instructions in visual-companion.md: Claude Code (default mode), Codex (auto-foreground via `CODEX_CI`), Gemini CLI (`--foreground` with `is_background`), and fallback for other environments
- Server now writes startup JSON to `$SCREEN_DIR/.server-info` so agents can find the URL and port even when stdout is hidden by background execution
- Single quotes around `${CLAUDE_PLUGIN_ROOT}` in hooks.json fail on Windows (cmd.exe doesn't recognize single quotes as path delimiters) and on Linux (single quotes prevent variable expansion)
- Fix: replaced single quotes with escaped double quotes — works across macOS bash, Windows cmd.exe, Windows Git Bash, and Linux, with and without spaces in paths
- Verified on Windows 11 (NT 10.0.26200.0) with Claude Code 2.1.72 and Git for Windows
**Brainstorming spec review loop skipped** (#677)
- The spec review loop (dispatch spec-document-reviewer subagent, iterate until approved) existed in the prose "After the Design" section but was missing from the checklist and process flow diagram
- Since agents follow the diagram and checklist more reliably than prose, the spec review step was being skipped entirely
- Added step 7 (spec review loop) to the checklist and corresponding nodes to the dot graph
- Tested with `claude --plugin-dir` and `claude-session-driver`: worker now correctly dispatches the reviewer
**Cursor install command** (PR #676)
- Fixed Cursor install command in README: `/plugin-add` → `/add-plugin` (confirmed via Cursor 2.5 release announcement)
Writing-plans no longer offers a choice between subagent-driven and executing-plans. On harnesses with subagent support (Claude Code, Codex), subagent-driven-development is required. Executing-plans is reserved for harnesses without subagent capability, and now tells the user that Superpowers works better on a subagent-capable platform.
`/brainstorm`, `/write-plan`, and `/execute-plan` now show deprecation notices pointing users to the corresponding skills. Commands will be removed in the next major release.
Optional browser-based companion for brainstorming sessions. When a topic would benefit from visuals, the brainstorming skill offers to show mockups, diagrams, comparisons, and other content in a browser window alongside terminal conversation.
-`lib/brainstorm-server/` — WebSocket server with browser helper library, session management scripts, and dark/light themed frame template ("Superpowers Brainstorming" with GitHub link)
-`skills/brainstorming/visual-companion.md` — Progressive disclosure guide for server workflow, screen authoring, and feedback collection
- Brainstorming skill adds a visual companion decision point to its process flow: after exploring project context, the skill evaluates whether upcoming questions involve visual content and offers the companion in its own message
- Per-question decision: even after accepting, each question is evaluated for whether browser or terminal is more appropriate
- **Brainstorming** — New sections: "Design for isolation and clarity" (clear boundaries, well-defined interfaces, independently testable units) and "Working in existing codebases" (follow existing patterns, targeted improvements only)
- **Writing-plans** — New "File Structure" section: map out files and responsibilities before defining tasks. New "Scope Check" backstop: catch multi-subsystem specs that should have been decomposed during brainstorming
- **SDD implementer** — New "Code Organization" section (follow plan's file structure, report concerns about growing files) and "When You're in Over Your Head" escalation guidance
- **SDD code quality reviewer** — Now checks architecture, unit decomposition, plan conformance, and file growth
- **Spec/plan reviewers** — Architecture and file size added to review criteria
- **Scope assessment** — Brainstorming now assesses whether a project is too large for a single spec. Multi-subsystem requests are flagged early and decomposed into sub-projects, each with its own spec → plan → implementation cycle
- **Model selection** — Guidance for choosing model capability by task type: cheap models for mechanical implementation, standard for integration, capable for architecture and review
- **Implementer status protocol** — Subagents now report DONE, DONE_WITH_CONCERNS, BLOCKED, or NEEDS_CONTEXT. Controller handles each status appropriately: re-dispatching with more context, upgrading model capability, breaking tasks apart, or escalating to human
Added `<SUBAGENT-STOP>` block to using-superpowers. Subagents dispatched for specific tasks now skip the skill instead of activating the 1% rule and invoking full skill workflows.
Superpowers now works with Cursor's plugin system. Includes a `.cursor-plugin/plugin.json` manifest and Cursor-specific installation instructions in the README. The SessionStart hook output now includes an `additional_context` field alongside the existing `hookSpecificOutput.additionalContext` for Cursor hook compatibility.
This fix should dramatically improve superpowers skills compliance and should reduce the chances of Claude entering its native plan mode unintentionally.
### Changed
**Brainstorming skill now enforces its workflow instead of describing it**
Models were skipping the design phase and jumping straight to implementation skills like frontend-design, or collapsing the entire brainstorming process into a single text block. The skill now uses hard gates, a mandatory checklist, and a graphviz process flow to enforce compliance:
-`<HARD-GATE>`: no implementation skills, code, or scaffolding until design is presented and user approves
- Explicit checklist (6 items) that must be created as tasks and completed in order
- Graphviz process flow with `writing-plans` as the only valid terminal state
- Anti-pattern callout for "this is too simple to need a design" — the exact rationalization models use to skip the process
- Design section sizing based on section complexity, not project complexity
Added an `EnterPlanMode` intercept to the skill flow graph. When the model is about to enter Claude's native plan mode, it checks whether brainstorming has happened and routes through the brainstorming skill instead. Plan mode is never entered.
### Fixed
**SessionStart hook now runs synchronously**
Changed `async: true` to `async: false` in hooks.json. When async, the hook could fail to complete before the model's first turn, meaning using-superpowers instructions weren't in context for the first message.
The `superpowers-codex` bootstrap CLI, Windows `.cmd` wrapper, and related bootstrap content file have been removed. Codex now uses native skill discovery via `~/.agents/skills/superpowers/` symlink, so the old `use_skill`/`find_skills` CLI tools are no longer needed.
Claude Code 2.1.x changed how hooks execute on Windows: it now auto-detects `.sh` files in commands and prepends `bash`. This broke the polyglot wrapper pattern because `bash "run-hook.cmd" session-start.sh` tries to execute the `.cmd` file as a bash script.
Fix: hooks.json now calls session-start.sh directly. Claude Code 2.1.x handles the bash invocation automatically. Also added .gitattributes to enforce LF line endings for shell scripts (fixes CRLF issues on Windows checkout).
The synchronous SessionStart hook blocked the TUI from entering raw mode on Windows, freezing all keyboard input. Running the hook async prevents the freeze while still injecting superpowers context.
The character-by-character loop using `${input:$i:1}` was O(n^2) in bash due to substring copy overhead. On Windows Git Bash this took 60+ seconds. Replaced with bash parameter substitution (`${s//old/new}`) which runs each pattern as a single C-level pass — 7x faster on macOS, dramatically faster on Windows.
- Windows doesn't respect shebangs, so directly invoking the extensionless `superpowers-codex` script triggered an "Open with" dialog. All invocations now prefixed with `node`.
- Fixed `~/` path expansion on Windows — PowerShell doesn't expand `~` when passed as an argument to `node`. Changed to `$HOME` which expands correctly in both bash and PowerShell.
**Worktree isolation now required before implementation**
Added `using-git-worktrees` as a required skill for both `subagent-driven-development` and `executing-plans`. Implementation workflows now explicitly require setting up an isolated worktree before starting work, preventing accidental work directly on main.
**Main branch protection softened to require explicit consent**
Instead of prohibiting main branch work entirely, the skills now allow it with explicit user consent. More flexible while still ensuring users are aware of the implications.
**Simplified installation verification**
Removed `/help` command check and specific slash command list from verification steps. Skills are primarily invoked by describing what you want to do, not by running specific commands.
**Codex: Clarified subagent tool mapping in bootstrap**
**OpenCode: Standardized on `plugins/` directory per official docs (#343)**
OpenCode's official documentation uses `~/.config/opencode/plugins/` (plural). Our docs previously used `plugin/` (singular). While OpenCode accepts both forms, we've standardized on the official convention to avoid confusion.
Changes:
- Renamed `.opencode/plugin/` to `.opencode/plugins/` in repo structure
- Updated all installation docs (INSTALL.md, README.opencode.md) across all platforms
Superpowers for OpenCode now uses OpenCode's native `skill` tool instead of custom `use_skill`/`find_skills` tools. This is a cleaner integration that works with OpenCode's built-in skill discovery.
**Migration required:** Skills must be symlinked to `~/.config/opencode/skills/superpowers/` (see updated installation docs).
### Fixes
**OpenCode: Fixed agent reset on session start (#226)**
The previous bootstrap injection method using `session.prompt({ noReply: true })` caused OpenCode to reset the selected agent to "build" on first message. Now uses `experimental.chat.system.transform` hook which modifies the system prompt directly without side effects.
**OpenCode: Fixed Windows installation (#232)**
- Removed dependency on `skills-core.js` (eliminates broken relative imports when file is copied instead of symlinked)
- Added comprehensive Windows installation docs for cmd.exe, PowerShell, and Git Bash
- Documented proper symlink vs junction usage for each platform
**Claude Code: Fixed Windows hook execution for Claude Code 2.1.x**
Claude Code 2.1.x changed how hooks execute on Windows: it now auto-detects `.sh` files in commands and prepends `bash `. This broke the polyglot wrapper pattern because `bash "run-hook.cmd" session-start.sh` tries to execute the .cmd file as a bash script.
Fix: hooks.json now calls session-start.sh directly. Claude Code 2.1.x handles the bash invocation automatically. Also added .gitattributes to enforce LF line endings for shell scripts (fixes CRLF issues on Windows checkout).
**Strengthened using-superpowers skill for explicit skill requests**
Addressed a failure mode where Claude would skip invoking a skill even when the user explicitly requested it by name (e.g., "subagent-driven-development, please"). Claude would think "I know what that means" and start working directly instead of loading the skill.
Changes:
- Updated "The Rule" to say "Invoke relevant or requested skills" instead of "Check for skills" - emphasizing active invocation over passive checking
- Added "BEFORE any response or action" - the original wording only mentioned "response" but Claude would sometimes take action without responding first
- Added reassurance that invoking a wrong skill is okay - reduces hesitation
- Added new red flag: "I know what that means" → Knowing the concept ≠ using the skill
**Added explicit skill request tests**
New test suite in `tests/explicit-skill-requests/` that verifies Claude correctly invokes skills when users request them by name. Includes single-turn and multi-turn test scenarios.
Added `disable-model-invocation: true` to all three slash commands (`/brainstorm`, `/execute-plan`, `/write-plan`). Claude can no longer invoke these commands via the Skill tool—they're restricted to manual user invocation only.
The underlying skills (`superpowers:brainstorming`, `superpowers:executing-plans`, `superpowers:writing-plans`) remain available for Claude to invoke autonomously. This change prevents confusion when Claude would invoke a command that just redirects to a skill anyway.
Fixed a confusing pattern where Claude would invoke a skill via the Skill tool, then try to Read the skill file separately. The `using-superpowers` skill now explicitly states that the Skill tool loads skill content directly—no need to read files.
- Added "How to Access Skills" section to `using-superpowers`
- Changed "read the skill" → "invoke the skill" in instructions
- Updated slash commands to use fully qualified skill names (e.g., `superpowers:brainstorming`)
2.**Code quality review** - Only runs after spec compliance passes. Reviews for clean code, test coverage, maintainability.
This catches the common failure mode where code is well-written but doesn't match what was requested. Reviews are loops, not one-shot: if reviewer finds issues, implementer fixes them, then reviewer checks again.
Other subagent workflow improvements:
- Controller provides full task text to workers (not file references)
- Workers can ask clarifying questions before AND during work
- Self-review checklist before reporting completion
- Plan read once at start, extracted to TodoWrite
New prompt templates in `skills/subagent-driven-development/`:
-`implementer-prompt.md` - Includes self-review checklist, encourages questions
-`spec-reviewer-prompt.md` - Skeptical verification against requirements
-`code-quality-reviewer-prompt.md` - Standard code review
**Debugging techniques consolidated with tools**
`systematic-debugging` now bundles supporting techniques and tools:
-`root-cause-tracing.md` - Trace bugs backward through call stack
-`defense-in-depth.md` - Add validation at multiple layers
-`condition-based-waiting.md` - Replace arbitrary timeouts with condition polling
-`find-polluter.sh` - Bisection script to find which test creates pollution
-`condition-based-waiting-example.ts` - Complete implementation from real debugging session
**Testing anti-patterns reference**
`test-driven-development` now includes `testing-anti-patterns.md` covering:
- Testing mock behavior instead of real behavior
- Adding test-only methods to production classes
- Mocking without understanding dependencies
- Incomplete mocks that hide structural assumptions
**Skill test infrastructure**
Three new test frameworks for validating skill behavior:
`tests/skill-triggering/` - Validates skills trigger from naive prompts without explicit naming. Tests 6 skills to ensure descriptions alone are sufficient.
`tests/claude-code/` - Integration tests using `claude -p` for headless testing. Verifies skill usage via session transcript (JSONL) analysis. Includes `analyze-token-usage.py` for cost tracking.
`tests/subagent-driven-dev/` - End-to-end workflow validation with two complete test projects:
-`go-fractals/` - CLI tool with Sierpinski/Mandelbrot (10 tasks)
-`svelte-todo/` - CRUD app with localStorage and Playwright (12 tasks)
Rewrote key skills using DOT/GraphViz flowcharts as the authoritative process definition. Prose becomes supporting content.
**The Description Trap** (documented in `writing-skills`): Discovered that skill descriptions override flowchart content when descriptions contain workflow summaries. Claude follows the short description instead of reading the detailed flowchart. Fix: descriptions must be trigger-only ("Use when X") with no process details.
**Skill priority in using-superpowers**
When multiple skills apply, process skills (brainstorming, debugging) now explicitly come before implementation skills. "Build X" triggers brainstorming first, then domain skills.
**brainstorming trigger strengthened**
Description changed to imperative: "You MUST use this before any creative work—creating features, building components, adding functionality, or modifying behavior."
### Breaking Changes
**Skill consolidation** - Six standalone skills merged:
-`root-cause-tracing`, `defense-in-depth`, `condition-based-waiting` → bundled in `systematic-debugging/`
-`testing-skills-with-subagents` → bundled in `writing-skills/`
-`testing-anti-patterns` → bundled in `test-driven-development/`
- **render-graphs.js** - Tool to extract DOT diagrams from skills and render to SVG
- **Rationalizations table** in using-superpowers - Scannable format including new entries: "I need more context first", "Let me explore first", "This feels productive"
- **docs/testing.md** - Guide to testing skills with Claude Code integration tests
- Optimized superpowers bootstrap to eliminate redundant skill execution. The `using-superpowers` skill content is now provided directly in session context, with clear guidance to use the Skill tool only for other skills. This reduces overhead and prevents the confusing loop where agents would execute `using-superpowers` manually despite already having the content from session start.
- Simplified `brainstorming` skill to return to original conversational vision. Removed heavyweight 6-phase process with formal checklists in favor of natural dialogue: ask questions one at a time, then present design in 200-300 word sections with validation. Keeps documentation and implementation handoff features.
- Updated `brainstorming` skill to require autonomous recon before questioning, encourage recommendation-driven decisions, and prevent agents from delegating prioritization back to humans.
- Clarified `writing-skills` guidance so it points to the correct agent-specific personal skill directories (`~/.claude/skills` for Claude Code, `~/.codex/skills` for Codex).
- Updated step 3: "Use the Read tool" → "Use the Skill tool to read and run"
- Updated rationalization list: "Read the current version" → "Run the current version"
The Skill tool is the proper mechanism for invoking skills in Claude Code. This update corrects the bootstrap instructions to guide agents toward the correct tool.
### Files Changed
- Updated: `skills/using-superpowers/SKILL.md` - Changed tool references from Read to Skill
**Strengthened using-superpowers skill against agent rationalization**
- Added EXTREMELY-IMPORTANT block with absolute language about mandatory skill checking
- "If even 1% chance a skill applies, you MUST read it"
- "You do not have a choice. You cannot rationalize your way out."
- Added MANDATORY FIRST RESPONSE PROTOCOL checklist
- 5-step process agents must complete before any response
- Explicit "responding without this = failure" consequence
- Added Common Rationalizations section with 8 specific evasion patterns
- "This is just a simple question" → WRONG
- "I can check files quickly" → WRONG
- "Let me gather information first" → WRONG
- Plus 5 more common patterns observed in agent behavior
These changes address observed agent behavior where they rationalize around skill usage despite clear instructions. The forceful language and pre-emptive counter-arguments aim to make non-compliance harder.
### Files Changed
- Updated: `skills/using-superpowers/SKILL.md` - Added three layers of enforcement to prevent skill-skipping rationalization
- **Fixed command syntax in README** (#44) - Updated all command references to use correct namespaced syntax (`/superpowers:brainstorm` instead of `/brainstorm`). Plugin-provided commands are automatically namespaced by Claude Code to avoid conflicts between plugins.
- Added copyable workflow checklist for tracking progress
- Added decision flowchart for when to revisit earlier phases
- Added comprehensive AskUserQuestion tool guidance with concrete examples
- Added "Question Patterns" section explaining when to use structured vs open-ended questions
- Restructured Key Principles as scannable table
**Anthropic best practices integration**
- Added `skills/writing-skills/anthropic-best-practices.md` - Official Anthropic skill authoring guide
- Referenced in writing-skills SKILL.md for comprehensive guidance
- Provides patterns for progressive disclosure, workflows, and evaluation
### Improvements
**Skill cross-reference clarity**
- All skill references now use explicit requirement markers:
-`**REQUIRED BACKGROUND:**` - Prerequisites you must understand
-`**REQUIRED SUB-SKILL:**` - Skills that must be used in workflow
-`**Complementary skills:**` - Optional but helpful related skills
- Removed old path format (`skills/collaboration/X` → just `X`)
- Updated Integration sections with categorized relationships (Required vs Complementary)
- Updated cross-reference documentation with best practices
**Alignment with Anthropic best practices**
- Fixed description grammar and voice (fully third-person)
- Added Quick Reference tables for scanning
- Added workflow checklists Claude can copy and track
- Appropriate use of flowcharts for non-obvious decision points
- Improved scannable table formats
- All skills well under 500-line recommendation
### Bug Fixes
- **Re-added missing command redirects** - Restored `commands/brainstorm.md` and `commands/write-plan.md` that were accidentally removed in v3.0 migration
- Fixed `defense-in-depth` name mismatch (was `Defense-in-Depth-Validation`)
- Fixed `receiving-code-review` name mismatch (was `Code-Review-Reception`)
- Fixed `commands/brainstorm.md` reference to correct skill name
- Removed references to non-existent related skills
### Documentation
**writing-skills improvements**
- Updated cross-referencing guidance with explicit requirement markers
- Added reference to Anthropic's official best practices
- Improved examples showing proper skill reference format
- **Fixed false warning when local skills repo is ahead of upstream** - The initialization script was incorrectly warning "New skills available from upstream" when the local repository had commits ahead of upstream. The logic now correctly distinguishes between three git states: local behind (should update), local ahead (no warning), and diverged (should warn).
## v2.0.1 (2025-10-12)
### Bug Fixes
- **Fixed session-start hook execution in plugin context** (#8, PR #9) - The hook was failing silently with "Plugin hook error" preventing skills context from loading. Fixed by:
- Using `${BASH_SOURCE[0]:-$0}` fallback when BASH_SOURCE is unbound in Claude Code's execution context
- Adding `|| true` to handle empty grep results gracefully when filtering status flags
The headline change is **skills repository separation**: all skills, scripts, and documentation have moved from the plugin into a dedicated repository ([obra/superpowers-skills](https://github.com/obra/superpowers-skills)). This transforms superpowers from a monolithic plugin into a lightweight shim that manages a local clone of the skills repository. Skills auto-update on session start. Users fork and contribute improvements via standard git workflows. The skills library versions independently from the plugin.
Beyond infrastructure, this release adds nine new skills focused on problem-solving, research, and architecture. We rewrote the core **using-skills** documentation with imperative tone and clearer structure, making it easier for Claude to understand when and how to use skills. **find-skills** now outputs paths you can paste directly into the Read tool, eliminating friction in the skills discovery workflow.
Users experience seamless operation: the plugin handles cloning, forking, and updating automatically. Contributors find the new architecture makes improving and sharing skills trivial. This release lays the foundation for skills to evolve rapidly as a community resource.
**The biggest change:** Skills no longer live in the plugin. They've been moved to a separate repository at [obra/superpowers-skills](https://github.com/obra/superpowers-skills).
**What this means for you:**
- **First install:** Plugin automatically clones skills to `~/.config/superpowers/skills/`
- **Forking:** During setup, you'll be offered the option to fork the skills repo (if `gh` is installed)
- **Updates:** Skills auto-update on session start (fast-forward when possible)
- **Contributing:** Work on branches, commit locally, submit PRs to upstream
- **No more shadowing:** Old two-tier system (personal/core) replaced with single-repo branch workflow
**Migration:**
If you have an existing installation:
1. Your old `~/.config/superpowers/.git` will be backed up to `~/.config/superpowers/.git.bak`
2. Old skills will be backed up to `~/.config/superpowers/skills.bak`
3. Fresh clone of obra/superpowers-skills will be created at `~/.config/superpowers/skills/`
