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 Ashkaan/contextium --skill implement-auditgit clone --depth 1 https://github.com/Ashkaan/contextiumWrote 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/ashkaan/contextium/implement-audit)<a href="https://agentmods.dev/skills/ashkaan/contextium/implement-audit"><img src="https://agentmods.dev/badge/skills/ashkaan/contextium/implement-audit/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/ashkaan/contextium/implement-audit"><img src="https://agentmods.dev/badge/skills/ashkaan/contextium/implement-audit.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector warn
SkillSpector: 6 findings, up to medium
These are SkillSpector’s own severities. On a checked sample its high-severity flags on skills were ~96% false positives — a documented command, a public API, a “never do X” rule — so we show them as a caution to read, not a verdict. Why →
- medium Excessive Agency · line 44 Skill enables autonomous high-impact decisions without human-in-the-loop verification. Critical operations (destructive commands, financial transactions, data deletion) should require explicit user confirmation.Fix: Add human-in-the-loop confirmation for destructive, irreversible, or high-impact operations. Never auto-execute commands that modify files, send data, or alter system state.
- medium Excessive Agency · line 158 Skill enables autonomous high-impact decisions without human-in-the-loop verification. Critical operations (destructive commands, financial transactions, data deletion) should require explicit user confirmation.Fix: Add human-in-the-loop confirmation for destructive, irreversible, or high-impact operations. Never auto-execute commands that modify files, send data, or alter system state.
- medium Agent Snooping · line 10 Skill enumerates or reads other installed skills. Access to other skills' SKILL.md files or the skills directory reveals prompt instructions, capabilities, and secrets that should be invisible to peer skills.Fix: Remove all code or instructions that list or read other skills' files or directories. Skills should operate independently; cross-skill access is a privilege escalation.
- medium Agent Snooping · line 18 Skill enumerates or reads other installed skills. Access to other skills' SKILL.md files or the skills directory reveals prompt instructions, capabilities, and secrets that should be invisible to peer skills.Fix: Remove all code or instructions that list or read other skills' files or directories. Skills should operate independently; cross-skill access is a privilege escalation.
- medium Agent Snooping · line 11 Skill enumerates or reads other installed skills. Access to other skills' SKILL.md files or the skills directory reveals prompt instructions, capabilities, and secrets that should be invisible to peer skills.Fix: Remove all code or instructions that list or read other skills' files or directories. Skills should operate independently; cross-skill access is a privilege escalation.
- medium Agent Snooping · line 19 Skill enumerates or reads other installed skills. Access to other skills' SKILL.md files or the skills directory reveals prompt instructions, capabilities, and secrets that should be invisible to peer skills.Fix: Remove all code or instructions that list or read other skills' files or directories. Skills should operate independently; cross-skill access is a privilege escalation.
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 | $0.00064 | $0.06557 |
| Opus 5.5 | $0.00026 | $0.02623 |
| Sonnet 5.5 | $0.00013 | $0.01311 |
| Haiku 4.5 | $0.00006 | $0.00656 |
Grade A, and why
implement-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 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 — 260 lines — stays where its author put it; the contents beside it link to each section on GitHub.
implement-audit — the loop's code reviewer
Review code just built in this conversation. Be adversarial — assume things were missed and find them. Do not confirm the work is good; find what's wrong.
implement-audit is the single code reviewer everywhere: /implement fires it at phase-4.7, and it is standalone-callable as /implement-audit after any ad-hoc substantial change. The retired bespoke code-review agent and the old per-surface dispatches all consolidated here.
Critical — Scope
Only review work from this conversation. Do NOT flag unrelated uncommitted changes, review full git status, comment on other sessions' work, or pad findings with non-issues. One exception, per that rule: a security or data-loss defect read in passing (exposed credential, injection, destructive-by-default path) is still reported — with the out-of-scope verdict, so it never holds the patch.
The skill has two reasoning halves. Step 1 runs automated checks in the main conversation (they need Bash + the session's tool access). Step 2 runs the adversarial-reasoning half on a different vendor from the author — whichever one the adversarial-review policy row selects — a reviewer from the model family that wrote the code has a prior commitment to "this is fine because I wrote it" that fresh context alone does not remove. The two halves merge into a single numbered list of findings with triage verdicts.
Runs at most once per session. Step 0's dedupe guard (scripts/audit-dedupe.ts) makes a hand-typed /implement-audit after /implement phase-4.7 a no-op that re-emits the first run's trailer, not a second full review.
Expected State
/implement's Phase 4 layers (validate.ts --phase checks) should have already run. If implement-audit consistently finds 3+ issues, flag this as a process failure — the layers aren't catching what they should, and the fix for that must reach the cause.
step-0-session-dedupe
node --experimental-strip-types .agents/skills/implement-audit/scripts/audit-dedupe.ts status
What ships with it
4 files 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.
- yesterday Changed b042f1f67a03
- 12d ago Changed · +2 lines 0298f94a8f5e
- 15d ago Changed d90efd1cdc4e
- 16d ago Changed · +84 lines · -25 tokens per session efc6349c9e72
- 17d ago Changed · -33 lines · +18 tokens per session 28cc1ccdcf6f
- 1mo ago Changed · +1 lines 7a890e60d9a3
- 1mo ago First seen · 206 lines · 71 tokens per session scan A 96bec57f014b
implement-audit is a skill published in the GitHub repository Ashkaan/contextium (36 stars, last pushed yesterday), licensed MIT. It adds 64 tokens to every session and 6,557 once invoked, about $0.0003 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.
Other skills, from other repositories
red-team-review
Unified adversarial review, two modes. Strategic Matrix v4.3 (MTA-004) for arguments/plans/decisions: Priors → Rubric → Adversarial Lenses → SWOT/TOWS → MCDA → Blind Spot/Kill Switch → Summary. Deliverable Audit for built artefacts (assignments, reports, code outputs): isolated fresh-context hunt against the brief…
decision-journal
Unified decision lifecycle: Pre-decision logging, post-decision review, failure classification, and calibration tracking. Absorbs: post-mortem-engine.
plan-review-architecture
Architecture-dimension reviewer for written plans. Use when running plan-review or directly when an architectural review is wanted. Activate for keywords like "architecture review", "data flow", "failure modes", "rollback", "edge cases", "test matrix". Scores 5 sub-dimensions 0-10, produces ranked fixes. Always cite…
code-review-loop
Use when opening a PR for review or when receiving review feedback. Activate for keywords like "code review", "PR review", "request review", "review feedback", "address comments", "reviewer said". Covers both ends of the loop: preparing a reviewable PR and acting on feedback rigorously. Always engage with every…
example-skill
Performs thorough code reviews following best practices. Use when reviewing pull requests, checking code quality, or ensuring coding standards compliance.
code-review
Analyze diffs, suggest improvements, and review pull requests with actionable feedback.