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/adamw7/tools/doc-checkgit clone --depth 1 https://github.com/adamw7/toolsWhat 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 | $0.00017 | $0.00496 |
| Opus 5 | $0.00009 | $0.00248 |
| Sonnet 5 | $0.00003 | $0.00099 |
| Haiku 4.5 | $0.00002 | $0.00050 |
Grade A, and why
doc-check 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
Run this repository's documentation contract check the way
.github/workflows/maven.yml runs it, and act on the result.
Both commands, in this order — the first publishes the rule, the second uses it:
mvn -pl claude-code-enforcer -am install
mvn -N validate -DenforceClaudeMd
Running only the second checks the new documents with the previously installed
rules, which is how a rule change appears to pass when it does not. If you
changed nothing under claude-code-enforcer, the first command is still cheap
insurance against a stale local artifact.
Then:
- Everything passed — say so, and name how many rules reported a pass. A rule that never ran also never fails, so the count is the evidence.
- A rule failed — read the violation lines rather than the summary. Each one names the file and the value it objected to. Fix the document, not the rule: these rules encode a contract that was agreed before the document was written, and a rule bent to fit a document guards nothing afterwards. If the rule really is wrong, say so explicitly and stop, rather than editing it in passing.
- The build failed before the rules ran — a missing JDK 25, a Maven outside
3.9.x, or an unresolved
tools.claude-code-enforcerartifact. That is a toolchain problem; report it as such, because the documents were not checked.
Two failures are worth naming precisely, because their message points away from their cause:
contextBudgetonCLAUDE.mdmeans the file outgrew 32 KB. The fix is moving detail intoAGENTS.mdor a skill, never trimming the required sections.moduleMapConsistencyafter adding a module means the module is not mentioned in bothCLAUDE.mdandAGENTS.md— add it to the module map in each.
Load the doc-contract skill before editing any of the three documents, and the
enforcer-rules skill before touching a rule.
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 · 44 lines · 17 tokens per session scan A 254b4c5aec7d
doc-check is a command published in the GitHub repository adamw7/tools (11 stars, last pushed 4d ago), licensed MIT. It adds 17 tokens to every session and 496 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 2026-08-30.
Other commands, from other repositories
validate
You are the lead validation orchestrator for the Cobalt project. Your job is to prove that Cobalt implements every WhatsApp Web / Desktop / Mobile feature correctly, both at the source level (behavioral parity with the WA Web JS source) and at the observable level (the Nodes, WAM events and HTTP that Cobalt produces…
review
Review current changes for correctness, style, and potential issues.
refine-title
深度分析 Issue 或 PR 内容，并将其标题重构为 Conventional Commits 格式.
fit-analyze-codescan
分析 Code Scanning 告警并创建安全分析文档.
Review Code Structure
Audit class responsibilities, package boundaries, duplication, and rule compliance across Issues.
Fix Sonar Issues
Run local static analysis and fix findings with conservative, behavior-preserving changes.