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 LucasSantana-Dev/sharekit --skill adt-rag-driftgit clone --depth 1 https://github.com/LucasSantana-Dev/sharekitWrote 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/lucassantana-dev/sharekit/adt-rag-drift)<a href="https://agentmods.dev/skills/lucassantana-dev/sharekit/adt-rag-drift"><img src="https://agentmods.dev/badge/skills/lucassantana-dev/sharekit/adt-rag-drift.svg" alt="Measured on agentmods" 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 | $0.00022 | $0.01044 |
| Opus 5 | $0.00011 | $0.00522 |
| Sonnet 5 | $0.00004 | $0.00209 |
| Haiku 4.5 | $0.00002 | $0.00104 |
Grade B, and why
adt-rag-drift scanned grade B with 1 finding 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 6d 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.
Reads agent configuration directoriesmediumAgent snooping
.claude/, .codex/, .gemini/ hold keys, settings and other credentials a mod has no legitimate need for.
cat ~/.claude/rag-index/weekly.md | grep -A 30 "Stale chunks" How it starts
The opening of the file, as written. The whole thing — 150 lines — stays where its author put it; the contents beside it link to each section on GitHub.
RAG Drift
Drift occurs when the index contains chunks from files that have been deleted or modified since indexing. The retrieval system silently returns outdated or orphaned content without error. Detect and fix drift before it degrades quality.
What is drift?
Two types of drift corrupt the index: see references/drift-types.md for details.
Drift is silent. Retrieval still works, but returns wrong or outdated info.
Detect drift
BLOCKED if External HD unmounted — report.py embedder cache unreachable; query local sqlite instead.
Check stale count: Read the drift summary to determine action:
- <5 stale chunks: normal churn; can wait for next full rebuild
- 5–20 chunks: fix incrementally today
-
20 chunks: full rebuild recommended (faster than 20+ incremental ops)
Run the report script:
cd ~/.claude/rag-index
venv/bin/python report.py
Read the drift summary:
cat ~/.claude/rag-index/weekly.md | grep -A 30 "Stale chunks"
See references/report-examples.md for example output.
Done when: report shows stale count and recommended action based on thresholds above.
Fix missing chunks (deleted files)
For files that no longer exist, delete their orphaned chunks:
sqlite3 ~/.claude/rag-index/index.sqlite
In the sqlite prompt:
-- View chunks from the deleted file
SELECT COUNT(*), path FROM chunks
WHERE path LIKE '%old-skill/SKILL.md%'
GROUP BY path;
-- Delete them
DELETE FROM chunks
WHERE path LIKE '%old-skill/SKILL.md%';
-- Confirm deletion
SELECT COUNT(*) FROM chunks
WHERE path LIKE '%old-skill/SKILL.md%';
Exit sqlite: .exit
Quick delete (one line):
sqlite3 ~/.claude/rag-index/index.sqlite \
"DELETE FROM chunks WHERE path = '~/.claude/old-skill/SKILL.md';"
Done when: sqlite confirms 0 chunks for that path (query returns no rows).
Fix modified chunks (updated files)
Files that exist but have changed since indexing need incremental reindex:
cd ~/.claude/rag-index
venv/bin/python build.py --incremental ~/.claude/standards/old-auth.md
What ships with it
2 files 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.
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.
- 6d ago First seen · 150 lines · 22 tokens per session scan B b516a6006e84
adt-rag-drift is a skill published in the GitHub repository LucasSantana-Dev/sharekit (1 stars, last pushed 6d ago), licensed MIT. It adds 22 tokens to every session and 1,044 once invoked, about $0.0001 per session on Opus 5. A static security scan graded it B with 1 finding (reads agent configuration directories). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
qmd-mcp-skill
Use a local QMD knowledge base through UXC over MCP stdio, with daemon-backed session reuse and typed retrieval flows that avoid repeated model warmup and unnecessary query-expansion latency.
bedrock-rag
Build RAG on Amazon Bedrock Knowledge Bases — ingestion, chunking, embeddings, vector stores, Retrieve and RetrieveAndGenerate, citations, and Guardrails contextual grounding. Use when a chatbot must answer from a document corpus with sources.
google-adk-python
Build AI agents with Google Agent Development Kit (ADK) for Python. Use when creating multi-agent systems, tool-using agents, or orchestrating LLM workflows with Google Cloud.
Cursor rules for Next
Cursor rules for Next.js development with Tailwind CSS and TypeScript integration.
ai-product
Every product will be AI-powered. The question is whether you'll build it right or ship a demo that falls apart in production. This skill covers LLM integration patterns, RAG architecture, prompt engineering that scales, AI UX that users trust, and cost optimization that doesn't bankrupt you. Use when: keywords…
molecular-rag
Retrieve structurally similar compounds with known properties from ChEMBL/ZINC to ground predictions and inform optimization. Based on MolRAG (Xian 2025, ACL).