ad-drift

A read-only audit that compares project documentation with the code that actually exists. It checks files such as agent instructions, architecture notes, and architecture decision records—documents that explain important technical choices—for inconsistencies.

In plain words
What is it for?
Use it to check documented technologies, commands, quality gates, folder boundaries, coding patterns, and decision-record statuses against the repository.
Why use it?
Project documentation can become outdated as code changes. This audit identifies mismatches without changing files, so the team can decide whether to update the documentation or the implementation.

Skill for Claude CodeCodex

Install

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.

agentmods
npx agentmods add skills/corridortech/posecap/ad-drift
Any agent
npx skills add CorridorTech/PoseCap --skill ad-drift
Clone the repo
git clone --depth 1 https://github.com/CorridorTech/PoseCap

Made for: Claude Code, Codex.

Per session 67 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 1,589 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

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.

ModelPer sessionOnce invoked
Fable 5 $0.00067 $0.01589
Opus 5 $0.00034 $0.00794
Sonnet 5 $0.00013 $0.00318
Haiku 4.5 $0.00007 $0.00159

Measured 2d ago against content hash 3173ac52d8c8, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

ad-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 2d 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.

.agents/skills/ad-drift/SKILL.md · 76 lines

How it starts

The opening of the file, as written. The whole thing — 76 lines — stays where its author put it; the contents beside it link to each section on GitHub.

<background_information> Read-only. Produces a drift list comparing the repo's operational docs against what the code actually does. Writes nothing — the user decides whether to fix the spec or the code. </background_information>

Step 2 — run checks.

AGENTS.md drift (if present):

  • Stack — does the listed stack match package.json / pyproject.toml / Cargo.toml / go.mod / equivalent?
  • Setup/build/test commands — do they match package.json#scripts, Makefile, or pyproject.toml?
  • Quality gates — do referenced hook configs exist (.husky/, .pre-commit-config.yaml, .github/workflows/)?
  • Repository layout — do referenced directories exist?
  • Pre-approved commands — anything in the list missing from the toolchain?

ARCHITECTURE.md drift (if present):

  • Layers and boundaries — do the named directories exist? Sample 1–2 files per layer; are imports respecting the stated boundaries?
  • Patterns — sample one handler / one repository / one entry point. Do they follow the documented pattern?
  • No ## Active ADRs section — ARCHITECTURE.md must not duplicate the doc/adr/ directory index per ADR-0030 §2.

ADR drift (if doc/adr/ exists):

  • Numbering — gaps or duplicates in doc/adr/NNNN-*.md?
  • Status field — every ADR has one of proposed | accepted | deprecated | superseded by ADR-NNNN.
  • Superseded chains — every "superseded by ADR-NNNN" target exists.

Spec drift (if doc/specs/ exists; structural integrity only — does NOT deep-audit spec text against code, deferred per ADR-0011):

  • Numbering — gaps or duplicates in doc/specs/NNNN-*.md?
  • Status field — every spec has one of draft | accepted | shipped | superseded by SPEC-NNNN.
  • Superseded chains — every "superseded by SPEC-NNNN" target exists.
  • Reciprocity — every task with non-empty Spec ref points to a spec that exists; every accepted/shipped spec has at least one entry in its Related → Tasks list.
  • No checkbox UI — per ADR-0030 §1, Spec is decision-record (not tracking). Functional Requirements / Non-functional Requirements / Success Criteria must use plain bullets, not - [ ] checkboxes; implementation tracking lives in per-Spec tasks.
  • Status / task aggregate alignment — when every task referencing a spec is done, the spec's Status should be shipped. A spec with all tasks done but Status: accepted is drift between work-unit completion and feature-level claim.

Documentation discipline drift (WORKFLOW.md §2 / ADR-0008). Audit narrative documents — README.md, AGENTS.md / CLAUDE.md, ARCHITECTURE.md, DESIGN.md, and prose pages under doc/ that are not lifecycle-managed artifacts under doc/product/, doc/specs/, doc/adr/, or doc/tasks/:

  • Emoji — any present? Rule 3 forbids emoji anywhere (docs, code, comments, commits, skill outputs).
  • Dates / version stamps / DRAFT markers / changelog blocks in narrative documents — Rule 2 forbids these. Lifecycle-managed artifacts under doc/product/, doc/specs/, doc/adr/, and doc/tasks/ are exempt.
  • Business context first — does the first paragraph answer why the document exists, before what and how? Rule 4.
  • Scope duplication — does the document copy material that is canonically owned by another file? Rule 5 requires linking, not copying.
  • Speculation — phrases like "we might", "in the future", "could be added", or roadmaps without an ADR / task reference. Rule 1 forbids unfounded plans.

Source code (sample, not exhaustive — flag findings, not every match):

  • Orphan TODO / FIXME — Rule 7. A reference to a GitHub Issue or a doc/tasks/NNNN-*.md task file makes it not orphan.
  • Commented-out code blocks — Rule 7. Removed code lives in git history.

Single-responsibility drift (ADR-0030 / WORKFLOW §2 rules #9–#12):

  • Definition-layer tracking UI (Rule #9) — grep ^- \[ \] / ^- \[x\] inside AGENTS.md, WORKFLOW.md, ARCHITECTURE.md, GUIDELINES.md, CONTEXT.md, doc/product/*.md. Definition documents must not carry per-item checkbox UI. Fenced code-block examples (PR-body templates, etc.) are illustrative, not pillar tracking.
  • Directory-as-index duplication (Rule #10) — flag sections that re-state another layer's index: ## Active ADRs inside ARCHITECTURE.md or AGENTS.md; multi-bullet ## Architectural Principles digests paraphrasing each ADR; PRD ## Related → ADRs bullet lists enumerating the kit's ADR ledger.
  • Kit-state in WORKFLOW.md (Rule #12) — grep ADR-[0-9]{4} in WORKFLOW.md. Universal philosophy must not cite kit-specific ADR numbers (downstream installs lack doc/adr/). Literature citations and generic doc/adr/ references are allowed.
  • Cross-references that are decoration (Rule #11) — sample inline per ADR-NNNN refs in narrative documents and apply the load-bearing test: deletion leaves the surrounding statement intact → decoration; flag.

Read the full file on GitHub · 76 lines

Files

What ships with it

1 file beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.

Changes

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.

  1. 2d ago First seen · 76 lines · 67 tokens per session scan A 3173ac52d8c8

Subscribe to this mod's changes

ad-drift is a skill published in the GitHub repository CorridorTech/PoseCap (190 stars, last pushed 10d ago), licensed Apache-2.0. It adds 67 tokens to every session and 1,589 once invoked, about $0.0003 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 2026-08-30.

Related

Other skills, from other repositories

systematic-debugging

Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes.

obra/superpowers · 21 tokens

brainstorming

You MUST use this before any creative work - creating features, building components, adding functionality, or modifying behavior. Explores user intent, requirements and design before implementation.

obra/superpowers · 37 tokens

chat-pet-sprite-creation

Use when creating or changing VS Code chat pet sprite art, sprite sheets, state animations, eye treatments, Stable/Insiders variants, or pet transitions under src/vs/workbench/contrib/chat/browser/widget/media/chatPet.

microsoft/vscode · 53 tokens

cpu-profile-analysis

Analyze V8/Chrome CPU profiles (.cpuprofile) and DevTools trace files (Trace-.json). Use when: profiling performance, investigating slow functions, comparing code paths, finding bottlenecks, analyzing timeToRequest, understanding call trees from sampling profiler data, analyzing layout/paint/rendering, investigating…

microsoft/vscode · 71 tokens

babysit-pr

Babysit a GitHub pull request after creation by continuously polling review comments, CI checks/workflow runs, and mergeability state until the PR is merged/closed or user help is required. Diagnose failures, retry likely flaky failures up to 3 times, auto-fix/push branch-related issues when appropriate, and keep…

openai/codex · 114 tokens

imagegen

Generate or edit raster images when the task benefits from AI-created bitmap visuals such as photos, illustrations, textures, sprites, mockups, or transparent-background cutouts. Use when Codex should create a brand-new image, transform an existing image, or derive visual variants from references, and the output…

openai/codex · 113 tokens