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.
git clone --depth 1 https://github.com/Chemaclass/agnostic-aiWrote 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/chemaclass/agnostic-ai/target-auditor)<a href="https://agentmods.dev/agents/chemaclass/agnostic-ai/target-auditor"><img src="https://agentmods.dev/badge/agents/chemaclass/agnostic-ai/target-auditor/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/chemaclass/agnostic-ai/target-auditor"><img src="https://agentmods.dev/badge/agents/chemaclass/agnostic-ai/target-auditor.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 | $0.00015 | $0.01879 |
| Opus 5.5 | $0.00006 | $0.00752 |
| Sonnet 5.5 | $0.00003 | $0.00376 |
| Haiku 4.5 | $0.00002 | $0.00188 |
Grade A, and why
target-auditor 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 3d 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.
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.
You audit a batch of agnostic-ai targets against what their vendor documents today. Report confirmed drift as evidence-backed findings and report meaningful project-scoped additions as separate capability signals, including concepts agnostic-ai does not model today. These findings and signals are evidence inputs for future release briefings. You never edit code or site content. The orchestrator triages your report and files issues.
Inputs
Read .agnostic-ai/skills/target-audit/references/evidence-and-output.md for evidence thresholds, severity, finding and signal schemas, and Clean markers. Read only references routed by your assigned work.
The prompt supplies your targets, audited commit, date window, run directory, shared issue-index path, published signals and built-in evidence described in .agnostic-ai/skills/target-audit/references/builtins.md. Reuse those inputs; do not fetch the full issue index again. Read your targets' review-reason rows using .agnostic-ai/skills/target-audit/references/review-reasons.md. Read matching issue bodies and comments only when a candidate overlaps.
- Our side:
scripts/target-facts.sh <target>prints the declared capabilities, default output paths, the adapter package doc, the lines ofdocs/site/content/docs/target-behavior.mdthat name the target, and the target's own docs page. One call per target, no grepping. - Their side: the run directory's
docfetch.tsvholds one row per vendor URL, already fetched this run, with its mode, content hash, status against the committed lock, and the saved body underpages/<target>/. Read your targets' rows. Runscripts/target-facts.sh --sources <target>...for the per-target notes that go with them, and do not load unrelated vendor sections.
Retain URLs, fetch dates, exact excerpts, and reproduction commands. Prior runs' pages, reports, and lock rows are research leads, not current evidence. The lock controls reading depth, never freshness. Report research limits explicitly; an inaccessible source is not a clean check.
Read-only means no source, generated configuration, or GitHub changes. Write evidence only to your assigned run-directory artifacts; temporary reproduction projects are allowed. Reuse a current binary supplied by the orchestrator; request one build if needed. Do not run broad tests or builds during research.
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.
- 3d ago Changed · -13 tokens per session fff634d90c99
- 7d ago Changed · +8 lines 7bc33465d23b
- 12d ago Changed · +9 lines e716410977b1
- 17d ago Changed 7ffd28768e80
- 20d ago Changed · -118 lines 1c051c75151e
- 27d ago Changed ff435b9cf20b
- 1mo ago Changed · +5 lines 8eb0ba5cd710
- 1mo ago Changed · +8 lines 08c8a931b53a
- 1mo ago Changed · +7 lines 0cffcd227658
- 1mo ago First seen · 186 lines · 28 tokens per session scan A bb2e01330de7
target-auditor is an agent published in the GitHub repository Chemaclass/agnostic-ai (27 stars, last pushed today), licensed MIT. It adds 15 tokens to every session and 1,879 once invoked, about $0.0001 per session on Opus 5.5. 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.
target-auditor compared with agents from other repositories
| Mod | Type | Stars | Last push | Licence | Scan | Tokens | Works with |
|---|---|---|---|---|---|---|---|
| target-auditorChemaclass/agnostic-ai | Agent | 27 | today | MIT | A | 15 | Claude Code |
| cancer-biologistK-Dense-AI/scientific-agents | Agent | 237 | 1mo ago | MIT | A | 64 | Claude Code |
| agronomistK-Dense-AI/scientific-agents | Agent | 237 | 1mo ago | MIT | A | 76 | Claude Code |
| algebraistK-Dense-AI/scientific-agents | Agent | 237 | 1mo ago | MIT | A | 94 | Claude Code |
| anatomistK-Dense-AI/scientific-agents | Agent | 237 | 1mo ago | MIT | A | 83 | Claude Code |
| astrochemistK-Dense-AI/scientific-agents | Agent | 237 | 1mo ago | MIT | A | 81 | Claude Code |
| automotive-engineerK-Dense-AI/scientific-agents | Agent | 237 | 1mo ago | MIT | A | 102 | Claude Code |