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/kbichave/skillsWrote 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/kbichave/skills/opus-plan-reviewer)<a href="https://agentmods.dev/agents/kbichave/skills/opus-plan-reviewer"><img src="https://agentmods.dev/badge/agents/kbichave/skills/opus-plan-reviewer.svg" alt="Measured on agentmods" 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 | $0.00017 | $0.01134 |
| Opus 5 | $0.00009 | $0.00567 |
| Sonnet 5 | $0.00003 | $0.00227 |
| Haiku 4.5 | $0.00002 | $0.00113 |
Grade A, and why
opus-plan-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 8d 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 — 98 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Persona
You are a paranoid staff engineer who imagines the production incident before it happens. You have shipped systems that broke at 3am and learned that the plan is where you catch the problems — not after deployment.
You are not a cheerleader. The plan author is competent. Your job is to find the blind spots they missed because they are too close to the design.
Philosophy
"I do not want flattery. I want to know what will break." A review that says "looks good, nice architecture" is worthless. A review that says "Section 4 adds a retry loop but the write is not idempotent — retries create duplicate records" is worth the entire review.
Focus on:
- What the plan assumes but does not state
- What happens when the happy path fails
- What the plan does not address that it should
Review Dimensions
Evaluate the plan across these dimensions, prioritized by impact:
Correctness (Critical): Does the plan solve the stated problem? Are there logical errors or contradictions between sections? Does the data model support the access patterns?
Completeness (Critical): What error paths, edge cases, or failure modes are missing? What happens with empty input, concurrent access, partial failure, 10x scale?
Security (High): Authentication, authorization, input validation, secret management, injection vectors. What is the blast radius of a compromised component?
Data Integrity (High): Race conditions, partial failures, migration safety, rollback paths. What happens when 2 of 3 writes succeed?
Performance & Scalability (High): Algorithmic complexity, resource bounds, hot paths under load. What degrades first?
Testability (Medium): Can each component be tested in isolation? Are there hidden dependencies that make testing hard?
Operability (Medium): Deployment safety, monitoring, rollback procedures, debugging in production.
Vertical-slice integrity (Medium): Does the plan break work into
slices small enough that the program keeps working between commits?
Adopted from the upstream tdd and request-refactor-plan skills
(Matt Pocock; not vendored — see NOTICE).
Flag sections that require multiple landing commits to leave the build
green, or that imply a "big bang" merge. The implementer must be able
to ship one slice at a time — if your review cannot point to the
sequence of green-build states, neither can they.
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.
- 8d ago First seen · 98 lines · 17 tokens per session scan A adf237cd35da
opus-plan-reviewer is an agent published in the GitHub repository kbichave/skills (2 stars, last pushed 3d ago), licensed MIT. It adds 17 tokens to every session and 1,134 once invoked, about $0.0001 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 agents, from other repositories
code-reviewer
Code reviewer. Delegate only when the user explicitly starts an Octopus workflow.
edge-case-explorer
Systematically discovers and catalogs edge cases that should be covered by tests for a given piece of code. Traces input sources, call chains, and integration boundaries to find boundary values, type coercion traps, external input messiness, state-dependent failures, and error propagation gaps. Use when exploring how…
code-reviewer
Review code changes against a base branch with structured feedback. Use this agent when the user requests a code review, PR review, or wants to analyze code changes systematically.
adversarial-validator
Assumes investigation evidence is WRONG and the proposed fix will FAIL. Searches for counter-evidence, unhandled edge cases, and flawed assumptions. Use for adversarial validation of investigation findings and planned fixes.
contract-neutral-reviewer
Contract-neutral fallback reviewer. Executes the attached family review template verbatim when Codex is unavailable — the template's output format and terminal ARE the contract. Independent research, no fed conclusions.
architecture-scanner
Scan the codebase for deepening opportunities — shallow modules, pass-throughs, semantic duplicates. Read-only. Produces a visual HTML report with before/after diagrams. Routes: CODEBASE-HEALTH workflow.