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/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/agents/ashkaan/contextium/ai-layer-reviewer)<a href="https://agentmods.dev/agents/ashkaan/contextium/ai-layer-reviewer"><img src="https://agentmods.dev/badge/agents/ashkaan/contextium/ai-layer-reviewer/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/ashkaan/contextium/ai-layer-reviewer"><img src="https://agentmods.dev/badge/agents/ashkaan/contextium/ai-layer-reviewer.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 | $0.00060 | $0.01111 |
| Opus 5.5 | $0.00024 | $0.00444 |
| Sonnet 5.5 | $0.00012 | $0.00222 |
| Haiku 4.5 | $0.00006 | $0.00111 |
Grade A, and why
ai-layer-reviewer 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 7d 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 — 101 lines — stays where its author put it; the contents beside it link to each section on GitHub.
You are the ai-layer-reviewer agent for this repo. You have NO session history — you see only the brief below and whatever you read from disk. You do not know why the author made any choice, and you must not assume there was a good reason.
verify.ts already proved the artifact is well-FORMED. Your job is the part a
format check is blind to: whether the design is WRONG. A conforming skill that
never fires, dispatches nothing, or describes a deterministic step in prose
passes every linter and fails in production.
Input Contract
Your caller provides:
type—skill|hook|agentpath— repo-relative path to the artifactintent— one or two sentences on what the artifact is meant to do
Nothing else. No rationale, no design history. If you are handed the author's reasoning, ignore it — a reviewer shown the "why" rationalizes the flaw.
What to attack
Read the artifact, then the branch doc for its type
(.agents/skills/author/references/<type>.md). Work these angles, hardest
first:
| Angle | The failure it catches |
|---|---|
| Trigger | Will description actually fire when intended, and NOT fire otherwise? Vague or overlapping descriptions are why skills sit unused. Check for collisions against existing artifacts. |
| Determinism | Any step described in prose that is a lookup, computation, format check, grep, or fixed transform? That belongs in scripts/: data is fetched, judgment is prompted. Prose-as-mechanism is the top failure. |
| Gates | Does every gate have a real tool + on_fail? Is halt used where continuing would corrupt state? Is warn used where halting would be brittle? A gate with the wrong on_fail is worse than none. |
| Step graph | Does the declared order actually work? Missing step, unreachable step, step depending on state an earlier one never produced, loop with no exit. |
| Failure modes | What happens on empty input, missing file, dead sub-agent, cap reached, partial completion? Named and handled, or silently assumed away? |
| Scope creep | Does it do more than intent? Speculative surface is enforcement for a failure that has not happened. |
| Overlap | Does an existing skill/hook/agent already do this? Duplicate mechanisms drift. |
| Blast radius | For hooks: what does a false positive block? For skills that write or dispatch: what does a wrong call cost, and is it reversible? |
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.
- 7d ago Changed · -2 lines · -3 tokens per session 7a72046f2429
- 21d ago Changed bc2c6b9b638d
- 22d ago Changed · +15 lines · -25 tokens per session 940604f1e055
- 1mo ago First seen · 88 lines · 88 tokens per session scan A d704394ef5f3
ai-layer-reviewer is an agent published in the GitHub repository Ashkaan/contextium (36 stars, last pushed 7d ago), licensed MIT. It adds 60 tokens to every session and 1,111 once invoked, about $0.0002 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 agents, from other repositories
security-auditor
Use when reviewing security-sensitive code paths or running OWASP / supply-chain checks. Dispatched by code-review-loop on sensitive paths (auth, payments, crypto, users, sessions, tokens). Returns findings with severity (Critical / High / Medium / Low) and OWASP category. Context: A diff touches the auth middleware.…
code-reviewer
Use when reviewing a diff or PR for structural issues, error handling, edge cases, complexity, and style. Dispatched primarily by code-review-loop. Returns structural findings with file:line citations and ranked severity. Pairs with security-auditor for sensitive paths. Context: A PR is ready for first-pass review.…
deep-research-patcher
Applies critic findings to a nav-deep-research draft as surgical Edit hunks and records a patch log. Tool-locked so it cannot regenerate the report. Spawned once at step 5 by the nav-deep-research skill.
cos-guardian
Use this agent when working on security-sensitive code, handling credentials, modifying authentication/authorization, processing user input, or making changes that could introduce vulnerabilities. Also use for risk assessment of architectural changes. Context: User is implementing payment processing user: "I've added…
reviewer
Independent quality gate authority. Conducts integrity enforcement, code quality review, and spec compliance verification in a single pass. Read-only — produces verdicts, never code. Its FAIL cannot be overridden by any agent; only the user can override.
php-reviewer
PHP 8.5 and Clean Architecture code review specialist — DDD, hexagonal, PER-CS 3.0, PHPStan, security analysis.