Borrowing it
Nothing to install: this file belongs to AHepi/DeepReason. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/AHepi/DeepReason/main/.claude/skills/dr-audit-spec-drift/SKILL.mdgit clone --depth 1 https://github.com/AHepi/DeepReasonWrote 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/ahepi/deepreason/dr-audit-spec-drift)<a href="https://agentmods.dev/skills/ahepi/deepreason/dr-audit-spec-drift"><img src="https://agentmods.dev/badge/skills/ahepi/deepreason/dr-audit-spec-drift/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/ahepi/deepreason/dr-audit-spec-drift"><img src="https://agentmods.dev/badge/skills/ahepi/deepreason/dr-audit-spec-drift.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
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 | $0.00053 | $0.00834 |
| Opus 5.5 | $0.00021 | $0.00334 |
| Sonnet 5.5 | $0.00011 | $0.00167 |
| Haiku 4.5 | $0.00005 | $0.00083 |
Grade A, and why
dr-audit-spec-drift 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 2mo 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 — 63 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Audit: code vs spec
Entry: LEDGER.md exists, spec-drift.md missing. Exit:
spec-drift.md written, LEDGER rows added, proofs in proof/.
Two directions, both mechanical:
- SPEC→TREE: a term the spec names that the tree no longer honors = the code deviates from spec.
- TREE→SPEC: a shipped public surface the spec never mentions = the spec is behind the code.
The spec series is every docs/harness-spec-*.md (base + all
amendments — read ALL amendments; later files supersede earlier ones
on conflict, so verify a term against the newest file naming it).
Operations
- SPEC term census:
rg -o -N '\x60[A-Za-z_][A-Za-z0-9_.:-]*\x60' docs/harness-spec-*.md | sort -u > proof/spec-terms.txtplus every##-level heading noun phrase. This is the SPEC list. - For each SPEC term:
rg -l -w '<term>' src/deepreason/(identifiers) orrg -l '<term>' src/ docs/map/(concept phrases). Zero hits → verdictspec-orphan(spec names it, tree does not have it), one row each with the scan saved. Hits → tally only. - TREE surface census, three lists with proofs:
a. CLI flags:
rg -o -N '"--[a-z][a-z0-9-]*"' src/deepreason/cli/main.py | sort -ub. Config fields:rg -o -N '^ [A-Z][A-Z0-9_]*:' src/deepreason/config.py | sort -uc. Typed error/refusal strings:rg -o -N '"[A-Z][A-Z0-9_]{6,}"' src/deepreason/run_manifest.py src/deepreason/preparation.py | sort -u - For each TREE surface item:
rg -l -F '<item>' docs/harness-spec-*.md. Zero hits → verdictspec-silent(shipped, undocumented in spec), one row each. Batch rows by feature, not per flag, when one tranche shipped them together (the row names the batch and lists members). - Write
spec-drift.md: two tables (one per direction) with count lines. Thespec-silenttable's parked prompt is a spec-amendment draft request (append-only amendment file, never an edit to existing spec text), route dr-change-orchestrator.
GATE
Pass: proofs exist for both censuses; every spec-orphan and every
spec-silent row cites its scan file; count lines match row counts.
Verdict labels: spec-orphan | spec-silent | covered.
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.
- 2mo ago First seen · 63 lines · 53 tokens per session scan A 51018ff5f27e
dr-audit-spec-drift is a skill published in the GitHub repository AHepi/DeepReason (141 stars, last pushed 1mo ago), licensed MIT. It adds 53 tokens to every session and 834 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 skills, from other repositories
autoreview
Structured code review through the shared OpenClaw agent-skills installation.
omh-code-review
This is an OMH code-review workflow skill, projected for Agent Skills hosts (Claude Code, Codex, Cursor, opencode, OpenClaw, pi).
revdiff-plan
Review the last Codex assistant message (plan, analysis, or proposal) with inline annotations in a TUI overlay. Extracts the most recent response from Codex rollout files and opens it in revdiff for review and annotation. Activates on "revdiff-plan", "review plan with revdiff", "annotate plan", "review last response"…
explore-unknowns
Guide the user through a quadrant walk that maps the unknowns of a task — open by listing the known knowns, then work through known unknowns, unknown knowns, and unknown unknowns one stage at a time, ending with a complete four-quadrant map in the user's hands. Use when a request is ambiguous or underspecified, the…
re0-merge
Review and land an external contribution the way this suite does: gate it against the thesis, land it with the author's credit intact, complete a new skill rather than merging it raw, then approve, credit, and explain before closing. Use when reviewing a pull request, as any collaborator or maintainer, not only the…
asplos-writing-style
Use when drafting or revising ASPLOS prose — engineering the first two pages that rapid review actually reads, stating the cross-layer insight as one quotable sentence, structuring per-layer mechanism sections, writing bounded claims, and fitting everything into 11 self-contained pages that include figures and…