Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add commands/end1989/repos-expert/refreshgit clone --depth 1 https://github.com/end1989/repos-expertWhat it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00011 | $0.00456 |
| Opus 5 | $0.00005 | $0.00228 |
| Sonnet 5 | $0.00002 | $0.00091 |
| Haiku 4.5 | $0.00001 | $0.00046 |
Grade A, and why
refresh scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
What it actually says
Bring the knowledge base back in line with the code.
1. Survey. For every directory in this folder containing a .git:
git -C <repo> rev-parse HEAD
Compare against _knowledge/<repo>/meta.json's sha. Sort every repo into:
- fresh — SHAs match. Nothing to do.
- stale — SHAs differ. The code moved since the docs were written.
- undocumented — no
_knowledge/<repo>/at all.
2. Report before spending anything. Print the three lists with counts, and for stale repos show how far behind they are:
git -C <repo> log --oneline <meta.sha>..HEAD | wc -l
Studying a repo takes a few minutes of model time each. Say how many are involved. If the
user passed --dry-run, stop here — that is the whole point of the flag.
3. Ask before a large batch. If more than 5 repos would be studied, ask first and wait. Below that, go ahead.
4. Restudy the stale ones using the same procedure as /study: gather git context,
launch curator subagents concurrently, accept only the five allowed filenames, write
meta.json with the SHA you read in step 1.
Pass the previous docs and the <meta.sha>..HEAD change log to each curator and tell it to
update, not rewrite — the point of a refresh is to revise what moved, not to lose
insight that is still true.
Do not automatically study undocumented repos. A folder can gain a repo without anyone
intending a multi-minute study of it. List them and let the user ask for /study <name>.
5. Offer the map. If anything changed, cross-repo-map.md may now be wrong too —
offer /map, don't run it unprompted.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 2d ago First seen · 43 lines · 11 tokens per session scan A ad6d0f8ae6f6
refresh is a command published in the GitHub repository end1989/repos-expert (0 stars, last pushed 2d ago), licensed MIT. It adds 11 tokens to every session and 456 once invoked, about $0.0001 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other commands, from other repositories
rb-setup
First-time setup. Configure the LLM API key, or a no-API-key local host runner (Codex / Trae / Claude / any headless CLI) that RepoBrain uses for codebase Q&A and refresh. / 首次 setup，配置 RepoBrain 代码问答与 refresh 所需的 LLM API key，或无需 API key 的本地 host runner（Codex / Trae / Claude / 任意无头 CLI）。.
wiki-query
Ask questions against the wiki. Synthesizes answers from wiki pages with cross-reference citations.
wiki-init
Initialize a new LLM Wiki in the current directory. Creates the full directory structure, config, and template files.
rp-reminder-cli
Continue your current workflow using rp-cli instead of built-in alternatives.
contribute
Add a thought, source, or paste. Use --project/-p to target another project without switching current. Usage: /mb:contribute [-p ] [--mode daily|concept|daily+concept].
build
Rebuild context.md for a project from sources/. Usage: /mb:build [-p ] [project-id].