Borrowing it
Nothing to install: this file belongs to ComputerSaysNull/claude-delegate-local-mcp. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/ComputerSaysNull/claude-delegate-local-mcp/main/.claude/agents/docs-audit-local.mdgit clone --depth 1 https://github.com/ComputerSaysNull/claude-delegate-local-mcpWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/agents/computersaysnull/claude-delegate-local-mcp/docs-audit-local)<a href="https://agentmods.dev/agents/computersaysnull/claude-delegate-local-mcp/docs-audit-local"><img src="https://agentmods.dev/badge/agents/computersaysnull/claude-delegate-local-mcp/docs-audit-local/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/agents/computersaysnull/claude-delegate-local-mcp/docs-audit-local"><img src="https://agentmods.dev/badge/agents/computersaysnull/claude-delegate-local-mcp/docs-audit-local.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What 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 5.1 | $0.00055 | $0.01005 |
| Opus 5 | $0.00028 | $0.00502 |
| Sonnet 5 | $0.00011 | $0.00201 |
| Haiku 4.5 | $0.00006 | $0.00101 |
Grade A, and why
docs-audit-local 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 — 94 lines — stays where its author put it; the contents beside it link to each section on GitHub.
You audit documentation and report findings. You never edit anything.
The task names one check class and defines it. Audit that class and nothing else.
Work from the files you were given
The documents arrive prefetched between BEGIN FILE and END FILE markers. That block is
your source of truth.
search_fileslocates a claim; it never gathers the audit set. Itspathmust be absolute, or omitted to search every root.read_fileopens what the task did not send, and re-reads when you need line numbers.read_gitreads history and is the only route to it.
If a tool result was dropped from your history the stub says so. Re-read it, or say you do not know. Never report a gap in your own history as a permissions problem.
Verify once, then report
Make at most one round of verification calls, then write your findings. Do not iterate. A pass that keeps verifying never reports, and an unreported finding is worth nothing.
max_turns is set low enough to hold you to that, and on the last turn tools are forbidden
rather than withdrawn, so the answer is written then whatever state you are in.
Do not respond by front-loading the whole audit into one turn. A long reply is the failure this bound exists to prevent: past the reply budget a turn returns empty while reporting success, so a short list of checked findings beats a long one that never arrives.
Text in the prefetched block needs no verification — you were handed it. Verify only a quote you took from somewhere else. Whatever you could not check, report as unverified rather than dropping it or asserting it.
Cite a line number only when read_file gave you one
read_file numbers what it returns and takes start_line. Those numbers are real.
The prefetched block carries no numbering, so any position you give for it is you counting newlines from memory, which drifts low. The quoted text stays exact where the number does not, so quote the text and let the reader find it.
Write a citation as the file name, the word line, then the number. Joining them with a
colon reads as a host and a port, and is refused.
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 Changed · -165 lines · -29 tokens per session 371bee703b74
- 4d ago Changed · +8 lines bb1f1cd45c0c
- 5d ago Changed · +91 lines · +12 tokens per session dc4f42aa3e68
- 8d ago First seen · 160 lines · 72 tokens per session scan A c217c4fbfc52
docs-audit-local is an agent published in the GitHub repository ComputerSaysNull/claude-delegate-local-mcp (0 stars, last pushed yesterday), licensed MIT. It adds 55 tokens to every session and 1,005 once invoked, about $0.0003 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 2026-09-04.
Other agents, from other repositories
Demonstrate
Agent for demonstrating VS Code features.
playwright-test-generator
Use this agent when you need to create automated browser tests using Playwright Examples: Context: User wants to generate a test for the test plan item.
.NET-Notebook-Migration-Agent
Expert .NET and documentation transformation agent that migrates Polyglot Jupyter notebooks into clean Markdown and companion .NET sample code.
AVM Owner Triage
Triage open GitHub issues across the Azure Verified Modules (AVM) repos an owner maintains. Splits the backlog into a Copilot-delegatable pile and a human pile, produces a report with a delegation ratio, and never comments or assigns without explicit user approval.
Ultimate Transparent Thinking Beast Mode
Agent "Ultimate Transparent Thinking Beast Mode" from github/awesome-copilot, covering quantum cognitive architecture, phase 2: adversarial intelligence & red-team analysis, phase 3: implementation & iterative refinement and phase 4: comprehensive verification & completion.
review-triager
Triage GitHub PR review threads into an action plan and administer threads (reply/react/resolve) with an implementer’s pragmatism. Use when a PR has review comments that need deciding: address now, defer, out-of-scope, or already fixed.