* docs(hermes): document plugging memory-tencentdb into an existing Hermes
The Hermes section in README.md / README_CN.md only documented the
all-in-one Docker image, which is the wrong starting point when the
user already has hermes-agent installed. They don't want to throw
away their setup just to get memory; they want to drop the provider
in next to the other built-in providers.
The full reference already exists at
`hermes-plugin/memory/memory_tencentdb/README.md` (auto-discovery
paths, env vars, supervisor behaviour, troubleshooting). What's
missing is the discoverable entry point at the top-level README.
Restructure section 2 into two parallel paths:
- 2.A Docker — the existing one-command flow, unchanged.
- 2.B Plug into an existing Hermes (no Docker) — new. Covers:
- installing the npm package so the Gateway source is on disk
- symlink vs copy of the Python provider into
`<hermes-agent-checkout>/plugins/memory/memory_tencentdb/`
(with the "underscore not hyphen" trap called out)
- the `memory.provider: memory_tencentdb` line in `~/.hermes/config.yaml`
- the three `MEMORY_TENCENTDB_LLM_*` env vars
- the three Gateway-start options (auto-discovery / explicit
GATEWAY_CMD / run-yourself), with auto-discovery as default
- a `curl /health` verification step
- a link to the in-repo provider README for the full reference
Docs-only — no code change. Existing Docker users see no difference.
Fixes#69
Signed-off-by: 李冠辰 <liguanchen@xiaomi.com>
* docs(hermes): document tdai gateway config file
* update README_CN.md file
* Fix formatting in README_CN.md for clarity
* Clarify Hermes plugin installation instructions
Updated the README to clarify installation instructions for the Hermes plugin, including Docker usage and configuration steps.
---------
Signed-off-by: 李冠辰 <liguanchen@xiaomi.com>
Co-authored-by: 十五便士 <95488710+Maxwell-Code07@users.noreply.github.com>
- Add `recall.maxCharsPerMemory` and `recall.maxTotalRecallChars` with defaults of `0`, which do not alter existing behavior. Users can opt in by setting positive values to cap injected memory context.
- Apply the budget after L1 search and before `<relevant-memories>` injection, preserving score order while truncating oversized entries and dropping overflow.
- Document the new guards in README, README_CN, and `openclaw.plugin.json`.
YOMXXX