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 fmind/dotfiles --skill diff-reviewgit clone --depth 1 https://github.com/fmind/dotfilesWrote 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/fmind/dotfiles/diff-review)<a href="https://agentmods.dev/skills/fmind/dotfiles/diff-review"><img src="https://agentmods.dev/badge/skills/fmind/dotfiles/diff-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/fmind/dotfiles/diff-review"><img src="https://agentmods.dev/badge/skills/fmind/dotfiles/diff-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 | $0.00057 | $0.01205 |
| Opus 5.5 | $0.00023 | $0.00482 |
| Sonnet 5.5 | $0.00011 | $0.00241 |
| Haiku 4.5 | $0.00006 | $0.00120 |
Grade A, and why
diff-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 — 65 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Diff Review
Find defects that would justify changing the candidate, with enough evidence for the author to reproduce and fix them.
Boundary
- Review only by default. Do not edit code, resolve threads, commit, push, or approve a pull request unless the user separately requests that action.
- Resolve the exact target and record base, head, working-tree state, and whether the evidence covers a dirty candidate, local commit, or remote pull-request head.
- Preserve staged, unstaged, and untracked work. Use an isolated temporary worktree when broad validation would mutate or misrepresent the candidate.
- Review against repository instructions, the issue or spec contract, and current source behavior. A green suite does not prove the intended behavior was implemented.
- Report only actionable defects, material test gaps, or explicit requested nits. Do not manufacture findings to make the review look useful.
Workflow
- Resolve intent and target: Read the request, issue, spec, plan, and change description. Identify the exact comparison and proof already supplied.
- Inventory the delta: Inspect changed files, generated artifacts, dependency or schema changes, and nearby code needed to understand behavior. Do not review the diff in isolation when invariants live elsewhere.
- Read tests first: Determine what behavior the candidate claims, whether tests can fail for the defect class, and which requirements remain unproved.
- Trace intended versus implemented: Map permissions, user journeys, data rules, failure semantics, and operational promises to concrete code paths and tests.
- Review by risk: Examine correctness, data integrity, authorization, input boundaries, concurrency, resource lifecycle, error propagation, compatibility, migration, performance, observability, and rollback in proportion to the change. Challenge pass-through abstractions, hidden dependency construction, tests that bypass the public seam, and hypothetical flexibility with no second concrete use.
- Verify candidates: Reproduce each suspected defect with code tracing, a focused test, or a safe isolated temporary experiment. Quote the specific file and line that makes the finding real.
- Run proportional checks: Start with focused tests and static analysis; run the repository-owned full gate when cost and target coherence permit. Record exactly which candidate each result covers.
- Calibrate: Discard preferences and speculation. Rank remaining findings by user impact, exploitability, data loss, regression likelihood, and confidence.
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 · 65 lines · 57 tokens per session scan A f7129ca24521
diff-review is a skill published in the GitHub repository fmind/dotfiles (4 stars, last pushed 1mo ago), licensed MIT. It adds 57 tokens to every session and 1,205 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 2026-08-31.
Other skills, from other repositories
ast-grep
Search, outline, and rewrite code structurally with ast-grep rules.
github-pull-request
Open, update, and verify GitHub pull requests.
repository-review
Read-only review of code, diffs, PRs, or repos for bugs and regressions.
code-security
Review code security; scan secrets and vulnerabilities with Gitleaks and Trivy.
full-review
Full review: read every repo area, apply verified fixes, and prove all checks pass.
duckdb
Query and export CSV, Parquet, JSON, or SQLite data with DuckDB.