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/sorah/configWrote 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/commands/sorah/config/review)<a href="https://agentmods.dev/commands/sorah/config/review"><img src="https://agentmods.dev/badge/commands/sorah/config/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/commands/sorah/config/review"><img src="https://agentmods.dev/badge/commands/sorah/config/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 | $0.00014 | $0.01470 |
| Opus 5.5 | $0.00006 | $0.00588 |
| Sonnet 5.5 | $0.00003 | $0.00294 |
| Haiku 4.5 | $0.00001 | $0.00147 |
Grade A, and why
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 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 — 123 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Conduct a specification interview session. Review a spec file, identify all ambiguities, gaps, and missing details, then interview the spec author about literally anything until the spec is complete and unambiguous. This is an in-depth, exhaustive process — do not stop early. Ask about every detail: naming, defaults, error messages, edge cases, field types, method signatures, return values, logging payloads, configuration keys, test expectations, and more. Continue until confident that an implementor could build the feature from the spec alone without further questions.
Input
The user provides a spec file path as argument: $ARGUMENTS
If no argument is given, ask the user which spec file to review.
Preparation Phase
Before asking any questions:
- Read the spec file thoroughly
- Read all referenced files — follow every file reference, schema file, existing code mentioned in the spec
- Read project conventions — check for project-level documentation (CLAUDE.md, style guides, architecture docs, knowledge files) relevant to the spec's domain
- Read project roadmap — look for big-picture documents, project timelines, roadmaps, or milestone plans that describe planned or upcoming changes relevant to the spec's domain
- Explore existing codebase for related models, modules, patterns, and conventions that the spec's implementation would interact with
- Check for inconsistencies — field collisions in schemas, index types, formula errors, duplicated sections, naming mismatches between spec and code
Interview Phase
Ask questions in focused batches of 2-4 using the AskUserQuestion tool. Group related questions together. After every batch of answers, update the spec file before continuing to the next batch. This keeps the spec current throughout the session and prevents losing context in long interviews.
Interview sessions are expected to be lengthy — often 10+ rounds of questions. Do not rush or skip areas. Cover these areas systematically, and revisit earlier areas if later answers reveal new ambiguities.
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 First seen · 123 lines · 14 tokens per session scan A 8e45ef13991d
review is a command published in the GitHub repository sorah/config (31 stars, last pushed yesterday), licensed MIT. It adds 14 tokens to every session and 1,470 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 2026-10-17.
Other commands, from other repositories
checklist
Generate a custom checklist for the current feature based on user requirements.
clarify
Identify underspecified areas in the current feature spec by asking up to 5 highly targeted clarification questions and encoding answers back into the spec.
specify
Create or update the feature specification from a natural language feature description.
converge
Assess the current codebase against the feature's spec, plan, and tasks, then append any remaining unbuilt work as new tasks to tasks.md so implement can complete it.
implement
Execute the implementation plan by processing and executing all tasks defined in tasks.md.
speckit.bug.fix
Apply the remediation from a bug assessment and record what was changed.