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 saajunaid/caddis-plugin --skill anchor-reviewgit clone --depth 1 https://github.com/saajunaid/caddis-pluginWrote 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/saajunaid/caddis-plugin/anchor-review)<a href="https://agentmods.dev/skills/saajunaid/caddis-plugin/anchor-review"><img src="https://agentmods.dev/badge/skills/saajunaid/caddis-plugin/anchor-review/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/saajunaid/caddis-plugin/anchor-review"><img src="https://agentmods.dev/badge/skills/saajunaid/caddis-plugin/anchor-review.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 | $0.00022 | $0.01179 |
| Opus 5.5 | $0.00009 | $0.00472 |
| Sonnet 5.5 | $0.00004 | $0.00236 |
| Haiku 4.5 | $0.00002 | $0.00118 |
Grade A, and why
anchor-review 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 1mo 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 — 136 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Anchor Review Skill
Purpose
Perform a rigorous adversarial review using a single model by systematically arguing both sides of every finding. This compensates for the lack of multi-model disagreement by forcing structured self-challenge.
Use this skill whenever you need high-confidence code review, fix verification, or implementation validation — especially for 🔴 findings, L-sized changes, hotfixes, or @anchor Evidence Bundles.
The 3-Lens Review
Review every changed file through three adversarial lenses, in order. Each lens has a different goal and a different failure mode it's designed to catch.
Lens 1: Correctness
"Does this code do what it claims to do?"
| Check | How to Verify |
|---|---|
| Logic matches spec/plan requirements | Cross-reference against plan/PRD |
| Edge cases handled (null, empty, zero, boundary) | Trace each input path mentally |
| Error paths return meaningful results | Read every except/catch block |
| State mutations are intentional | Check session state, database writes, file I/O |
| Return types match declared signatures | Verify type hints against actual returns |
Lens 2: Security
"Can this code be exploited or leak data?"
| Check | How to Verify |
|---|---|
| SQL injection | All queries use ? parameterization, no f-strings/format() |
| Secrets exposure | No hardcoded credentials; .env or pydantic-settings only |
| Input validation | User inputs validated before use (Pydantic, type checks) |
| PII in logs | logger.* calls don't include passwords, tokens, emails |
| Auth bypass | Protected routes check authentication before processing |
Lens 3: Performance
"Will this code break under real-world load?"
| Check | How to Verify |
|---|---|
| N+1 queries | Database calls inside loops → batch instead |
| Unbounded collections | Lists/dicts that grow without limit → add caps |
| Missing caching | Expensive repeated calls → @st.cache_data or equivalent |
| Large IN clauses | SQL IN with >900 params → batch to 900 |
| Blocking operations | Long I/O in main thread → async or background |
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.
- 1mo ago First seen · 136 lines · 22 tokens per session scan A dac669be836b
anchor-review is a skill published in the GitHub repository saajunaid/caddis-plugin (0 stars, last pushed today), licensed MIT. It adds 22 tokens to every session and 1,179 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 2026-09-03.
Other skills, from other repositories
goga-cookbook
Principles for applying DSL specification in cell and CODEMANIFEST design.
goga-review-plan
Verify execution plan completeness and correctness.
goga-define-challenge
Challenge the complete product definition before the PRD is generated.
goga-task-by-specifying
Interactive task formulation from a raw request.
goga-discover
Interview the user relentlessly about a decision until every branch of the design tree is resolved, then record the result as a short ADR.
goga-prototype-type-detail
Per-type detailing (methods, properties, signatures) for the prototype pipeline.