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/skymanbp/rag-your-code/describegit clone --depth 1 https://github.com/skymanbp/rag-your-codeWhat 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 | $0.00027 | $0.01059 |
| Opus 5 | $0.00014 | $0.00530 |
| Sonnet 5 | $0.00005 | $0.00212 |
| Haiku 4.5 | $0.00003 | $0.00106 |
Grade A, and why
describe 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 yesterday.
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.
How it starts
The opening of the file, as written. The whole thing — 105 lines — stays where its author put it; the contents beside it link to each section on GitHub.
/rag-your-code:describe
Write descriptions for declarations that have none. This is the work that turns an index into something worth searching, and nobody but a model that has read the code can do it.
Why this exists
Every unit's description is indexed. By default it is generated without a model: the identifier humanised, parameters and callees listed, the docstring appended. It introduces no vocabulary the source did not already have — which is exactly why a query for a concept nobody wrote down finds nothing.
Measured on this project's own rulers, moving from generated to agent-written descriptions took first-place accuracy from 0.314 to 0.429 and top-3 from 0.471 to 0.600. Against a Grep loop over the same questions it is the difference between 22.9% and 58.6% right-file-first.
The line this paragraph used to carry — that a cold index loses to Grep — came from one undescribed repository and did not survive being asked of two more. Cold, this side wins on Flask (37.1% to 22.9%), ties on cobra (17.5%), and lost on the retired subject. Which side wins undescribed depends on how much prose the repository already contains. What does not depend on it is the gain above.
What to do
-
See what is pending:
python -m ragyourcode.cli describe status -
Export a batch —
$ARGUMENTSunits if the user gave a number, else 20. Each entry carries the unit's source and a written brief:python -m ragyourcode.cli describe export --limit 20 > pending.json -
Read each unit's source and write its description. Follow the brief:
- say what the unit is for in domain terms — the operation, the failure it handles, the thing a person would search by;
- include the obvious synonyms for each, because those synonyms are the entire mechanism;
- do not restate the signature or list parameter names — that text is already indexed;
- do not describe behaviour the source does not show; the source is in the export so you can check;
- if a unit is trivial, say so briefly rather than padding it.
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.
- yesterday First seen · 105 lines · 27 tokens per session scan A a6af5caad796
describe is a command published in the GitHub repository skymanbp/rag-your-code (1 stars, last pushed 5d ago), licensed MIT. It adds 27 tokens to every session and 1,059 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
fieldtheory
Explain and drive the Field Theory CLI — bookmark-sourced seeds, repo-aware possibility runs that score ideas onto a 2x2 grid, and the interconnected .md files they leave behind. Trigger when the user asks about ft, bookmarks, seeds, ft possible (or the older name ft ideas), grids, dots/nodes, or how to turn saved…
rb-ask
Ask a question about the current project's codebase via the repobrain knowledge hub. / 通过 repobrain 知识库询问当前项目代码。.
rp-build-cli
Build with rp-cli context builder → chat → implement.
index
Index the codebase for semantic search.
add-agent
引导新增一个 Agent 适配器。用法 /add-agent.
index
Visual overview of Beacon index — chunks, coverage, provider. Pass --files to list indexed files.