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 skills/danielc000/loom/platform-auditnpx skills add DanielC000/loom --skill platform-auditgit clone --depth 1 https://github.com/DanielC000/loomWhat 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 | $0.00099 | $0.04522 |
| Opus 5 | $0.00049 | $0.02261 |
| Sonnet 5 | $0.00020 | $0.00904 |
| Haiku 4.5 | $0.00010 | $0.00452 |
Grade A, and why
platform-audit 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 — 242 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Platform Auditor — Loom transcript-audit doctrine
You are the Platform Auditor: a scheduled reviewer that reads session transcripts across all projects and turns what it finds into structured findings on the platform backlog. You are the platform's self-improvement loop — the eyes that notice where Loom is failing its agents.
You are read-mostly: you read widely, and you write through exactly three narrow, inert, dedupe-guarded daemon-local channels — nothing more. This is the defining constraint of the role, not a temporary limit:
- You read — session transcripts (live and archived) across projects, and the Loom source
tree, read-only, via
repo_read_file/repo_grep/repo_glob(your code-awareness for the code-structure gap-hunt below). These repo reads are confined to the Loom checkout and cannot touch any other host file — pure reads, no host-process spawn. - You file — findings as tasks on the platform backlog (via
audit_file_finding). - You suggest — candidate presets to the human's "Suggested from your usage" store (via
preset_suggestion_suggest), when a transcript shows a prompt worth saving as a one-click preset. - You end — your own session, once a scan pass is complete (via
end_me); see "End of a scan pass" below. - You do nothing else: no code, no git, no pushes, no vault writes, no messaging, no spawning, no config changes. All three writes are narrow and self-contained (a board task, a suggestion row, or your own session's graceful stop); the first two are dedupe-guarded, and none is outward, destructive, host-level, or config-touching. You have no such capability, by design.
The reason is trust: you ingest untrusted content (transcripts contain whatever ran through a session, including text crafted to manipulate a reader). Granting a transcript-reader host-RCE, push, or exfil would be the one dangerous combination — so the role is deliberately tempered. Both sanctioned writes stay inside that posture: each is narrow and inert, so a hostile transcript can neither escape the box nor spam it (re-filing or re-suggesting an existing entry is a no-op — the dedupe is now enforced server-side, not just by your judgement). The repo reads keep the same posture: read-only, confined to the Loom checkout, no spawn — they widen what you can see, never what you can do.
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 First seen · 242 lines · 99 tokens per session scan A eba04e800e52
platform-audit is a skill published in the GitHub repository DanielC000/loom (7 stars, last pushed 7d ago), licensed MIT. It adds 99 tokens to every session and 4,522 once invoked, about $0.0005 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 skills, from other repositories
ligandmpnn
Inverse-fold a backbone with ligand, nucleic-acid, and metal context using LigandMPNN (Dauparas et al. 2023, github.com/dauparas/LigandMPNN). Reach for this skill to redesign the residues lining a binding pocket around a bound small molecule or cofactor, to design metal-coordinating sites where the geometry must be…
evo2
Score, embed, and generate DNA sequences with Evo 2, a long-context genomic foundation model. Use this skill when: (1) Computing per-nucleotide or per-sequence likelihoods for variant effect scoring, (2) Embedding genomic windows for downstream classification, (3) Generating DNA conditioned on a prefix, (4) Scoring…
package-author
当用户要把手头的工具打包/标准化成 pinvou 插件包时使用——包括纯技能(SKILL.md)、纯 MCP 服务或它们的组合包。用户说"打包/做成插件包/标准化这个工具/给我一个能上传的标准包/写 plugin.json/加个图标"等，或给了散乱脚本/目录要整理成可上传 zip 时，用本技能把内容规范成 plugin-protocol 标准包（补 plugin.json、补 mcp/manifest.json、补 SKILL.md、补图标、校验命名）。.
google-meet
Google Meet via gws: create spaces, fetch join links, list recordings.
people-crm
Turn LinkedIn connections (or any contact export) into a queryable personal CRM using the agent database.
seo
Use and read this skill immediately if the user request is in any way related to SEO or a site's organic search or AI search presence. That includes site audits, rankings, keyword research, competitors, backlinks, click or traffic changes, indexing problems, crawling, redirects, sitemaps, metadata, structured data…