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 skills add cosai-oasis/secure-ai-tooling --skill audit-identification-questionsgit clone --depth 1 https://github.com/cosai-oasis/secure-ai-toolingWrote 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/skills/cosai-oasis/secure-ai-tooling/audit-identification-questions)<a href="https://agentmods.dev/skills/cosai-oasis/secure-ai-tooling/audit-identification-questions"><img src="https://agentmods.dev/badge/skills/cosai-oasis/secure-ai-tooling/audit-identification-questions/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/skills/cosai-oasis/secure-ai-tooling/audit-identification-questions"><img src="https://agentmods.dev/badge/skills/cosai-oasis/secure-ai-tooling/audit-identification-questions.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
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 | $0.00030 | $0.01153 |
| Opus 5 | $0.00015 | $0.00576 |
| Sonnet 5 | $0.00006 | $0.00231 |
| Haiku 4.5 | $0.00003 | $0.00115 |
Grade A, and why
audit-identification-questions 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 12d 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 — 97 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Identification Questions Audit
Audit persona identification questions for style guide compliance, boundary clarity, and term validity.
Scope
Target: a single persona id, or all to audit every non-deprecated persona in personas.yaml (default).
If a specific persona ID is provided (e.g., personaAgenticProvider), audit only that persona's questions.
Reference documents
Read these before auditing:
- Style guide (authoritative):
risk-map/docs/contributing/identification-questions-style-guide.md
Audit checklist
For each persona with identificationQuestions, verify ALL of the following. Report pass/fail per item.
Passing the mechanical checks (Structure, Examples format) does not make a question conformant. The editorial checks — Framing (an activity, not a role/title or a third party), Scoping, and especially "including" for scope expansion — take judgment against the style guide and are the most common remaining issue on a question that already looks clean. Do not return a conformant verdict until you have positively evaluated each editorial check on its merits — e.g. a base term narrow enough that a reader might wrongly exclude themselves, used without an "including" clause, is a scope-expansion flag even when structure and format are perfect.
Structure
- Count in range: read the count bound from the style guide's Rule enforcement summary (
validate-identification-questionshook) rather than restating the number here - Yes/no answerable: Every question can be answered with yes or no
- No embedded conditionals: No nested conditions requiring evaluation before answering
Framing
- Second-person, activity-focused: Questions ask what you do about the reader's own activities, using the approved second-person openers listed in the style guide's Rule enforcement summary. Avoid framings about a system/platform/product ("Is your system…", "Does your platform…") and third-person framings about other roles ("Do developers…").
- Activity-focused: Questions describe what the reader does, not what their job title is or what their product is
What ships with it
1 file beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
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.
- 12d ago First seen · 97 lines · 30 tokens per session scan A 1c8d72e85242
audit-identification-questions is a skill published in the GitHub repository cosai-oasis/secure-ai-tooling (97 stars, last pushed 5d ago), licensed Apache-2.0. It adds 30 tokens to every session and 1,153 once invoked, about $0.0002 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 skills, from other repositories
svc-mobile-android
Android APK static analysis — OWASP Mobile Top 10, Retrofit API audit, transport security, smali reading, component export, auth flow analysis. Use when target is an APK/Android app. Triggers - APK, Android, mobile app, decompiled, smali, jadx, apktool.
svc-pivoting
Turn a foothold into a RELIABLE pivot (SOCKS tunnel + persistent shell) so internal volume never rides a fragile stateless RCE. Use the moment you have code-exec on a dual-homed/edge host and need to reach an internal segment. Triggers - dual-homed host, "not reachable from my box", internal CIDR behind a foothold…
web-auth-bypass-idor
Broken access control - IDOR, privilege escalation, JWT abuse, mass assignment, forced browsing - for web apps and APIs. Use on any multi-user or role-based app with object IDs, tokens, or admin functionality. Triggers - /api/users/ , /orders/ , ?account=/uid=/doc=, JWT (eyJ...), role/isAdmin field, admin panel, "403…
web-deserialization
Insecure deserialization to RCE for web apps. Use when the app deserializes attacker-controlled data - cookies/tokens/hidden fields/params that are serialized blobs, VIEWSTATE, Java/PHP/.NET/Python/Node apps. Triggers - base64 starting rO0AB or hex AC ED 00 05 (Java), PHP serialize O-prefix, VIEWSTATE, python pickle…
web-sqli
SQL injection detection→exploitation→proof for web apps and APIs. Use when a parameter reaches a DB query, when input triggers a SQL error / boolean or time differential, or during VULN-ASSESSMENT/EXPLOITATION on a web target. Triggers - sql syntax error, ORA-/MySQL/psql/SQLite error string, ' or 1=1, order by…
web-upload-rce
File-upload abuse to code execution for web apps. Use when the app accepts a file (avatar, document, image, import, attachment) and you can influence name/type/content, or find where uploads land. Triggers - multipart upload form, avatar/profile picture, import/attachment, "invalid file type", uploaded file URL…