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 rjroy/vibe-garden --skill resolve-driftgit clone --depth 1 https://github.com/rjroy/vibe-gardenWrote 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/rjroy/vibe-garden/resolve-drift)<a href="https://agentmods.dev/skills/rjroy/vibe-garden/resolve-drift"><img src="https://agentmods.dev/badge/skills/rjroy/vibe-garden/resolve-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/rjroy/vibe-garden/resolve-drift"><img src="https://agentmods.dev/badge/skills/rjroy/vibe-garden/resolve-drift.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 | $0.00072 | $0.00517 |
| Opus 5 | $0.00036 | $0.00259 |
| Sonnet 5 | $0.00014 | $0.00103 |
| Haiku 4.5 | $0.00007 | $0.00052 |
Grade A, and why
resolve-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 10d 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 — 77 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Resolve Drift
Compare reference prose against living implementation evidence and reconcile drift.
This is the semantic pass. It is intentionally more expensive than lint.
Inputs
Read indexed pages from .lore/reference/index.md or .lore/reference/index.html.
For each page, read evidence metadata:
- Markdown:
fg-evidencefrontmatter - HTML:
fg-evidence-code,fg-evidence-tests, andfg-evidence-symbolsmeta tags
If evidence is missing, either:
- run
update-evidencefirst, or - audit manually and report that the page needs evidence
Audit Method
For each page:
- Extract concrete claims from the prose.
- Read the listed code and test evidence.
- Search nearby code with
rgwhen the evidence is incomplete. - Decide whether each claim is:
- current
- stale prose
- stale evidence
- implementation drift from intended design
- unverifiable policy or philosophy
Only flag contradictions when the code/tests and page make incompatible claims about the same subject. Do not treat different levels of detail as drift.
Parallel Review
For large wikis, split indexed pages into independent batches and use sub-agents when the user has allowed agent delegation. Each batch should return findings in this format:
Page:
Claim:
Evidence:
Verdict:
Suggested resolution:
Do not duplicate page batches across agents.
Fixing Drift
Prefer the smallest correction that restores truth:
- update stale reference prose when code/tests are clearly current
- update
fg-evidencewhen the prose is true but anchors are incomplete - report a code/test issue when the page documents intended design and implementation contradicts it
- mark historical pages as non-current when they are no longer meant to describe live behavior
Do not delete useful rationale just because implementation changed. Reframe it as history or a superseded decision when it still explains why the system evolved.
Output
Report grouped results:
- fixed drift
- remaining drift needing user decision
- evidence gaps
- pages audited with no actionable drift
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.
- 10d ago First seen · 77 lines · 72 tokens per session scan A 46f7455634dd
resolve-drift is a skill published in the GitHub repository rjroy/vibe-garden (6 stars, last pushed 2mo ago), licensed MIT. It adds 72 tokens to every session and 517 once invoked, about $0.0004 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 skills, from other repositories
procoder
Work like a senior developer in a repository governed by procoder: run the commit gate before calling anything done, format and lint through the binary, and drive the spec, plan, todo, backlog, and sprint chain in .procoder/. Use this skill when the repository contains a .procoder/ directory or an AGENTS.md naming…
code-review
The depth half of a review - the dimensions a diff is read against (correctness, boundaries, concurrency, failure paths, secrets, data access, structure, test quality) and the rule that a finding is refuted before it is reported. The verdict stays with the reviewer agent. Use when reviewing a diff or a pull request…
deslop
The optimization pass, defined - delete before you add, one smell class per pass, behaviour pinned by a test that ran BEFORE the edit. Lints a SKILL.md and prose by the same instinct. Use for the per-story optimization pass or when code has grown noisy without growing capable.
goal-plan-clean
A guide for creating a reusable plan-review loop. It repeatedly checks a plan, fixes blocking issues and suggestions, and stops after approval or a small number of review rounds.
loop-test
A local experiment for designing repeating checks with Claude Code’s loop command. It covers both fixed intervals and checks that schedule themselves again after each run.
take-task
A workflow for claiming a tracker task and preparing the development branch for it. It verifies the task, marks it as in progress, checks for uncommitted changes, and handles branch setup according to the task state.