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 agentmods add instructions/mick-gsk/drift/drift-agent-quickrefgit clone --depth 1 https://github.com/mick-gsk/driftWhat 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 | $0.01332 | $0.01332 |
| Opus 5 | $0.00666 | $0.00666 |
| Sonnet 5 | $0.00266 | $0.00266 |
| Haiku 4.5 | $0.00133 | $0.00133 |
Grade A, and why
drift drift-agent-quickref.instructions.md 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 3d 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 — 92 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Drift Agent Quickref
Diese Datei ist eine kompakte Einstiegshilfe. Sie ersetzt nicht die autoritativen Quellen:
POLICY.md.github/instructions/drift-policy.instructions.md.github/instructions/drift-push-gates.instructions.md
Task-Typ zu Pflicht-Checkliste
| Task-Typ | Aktive Gates | Pflichtdateien/Pflichtartefakte | Empfohlene Startbefehle |
|---|---|---|---|
feat: |
2, 3, 6, 7*, 8 | tests/**, benchmark_results/vX.Y.Z_*_feature_evidence.json, CHANGELOG.md, docs/STUDY.md, bei Signal/Ingestion/Output auch Audit-Artefakte |
make feat-start, danach make check |
fix: |
3, 6*, 7*, 8 | CHANGELOG.md, bei API-Docs Diff ggf. Docstring-Diff, bei Signal/Ingestion/Output Audit-Artefakte |
make fix-start, danach make test-fast |
| Signal-/Scoring-Arbeit | 2, 3, 7, 8 | Tests + Fixtures, Evidence, CHANGELOG.md, Audit-Artefakte (audit_results/*), ggf. ADR |
make feat-start, dann Risk-Audit aktualisieren |
chore: |
8 (plus ggf. 4/5 bei pyproject.toml) |
Bei Dependency-Aenderung immer konsistentes Lockfile | make check |
| Prompt/Instruction/Skill | Policy-Gate, ggf. 8 | Nur relevante Prompt-/Instruction-Dateien, keine Parallel-Policy | zuerst Policy-Gate, dann gezielter Edit |
| Review/Fix-Loop | Quality-Workflow + 8 | Review-Evidenz, konkrete Findings, anschliessende Re-Verification | make test-fast oder make check |
* Gate 6 gilt nur wenn src/drift/** betroffen ist. Gate 7 gilt nur bei Aenderungen an src/drift/signals/**, src/drift/ingestion/** oder src/drift/output/**.
Schnellregeln
- Vor jeder Umsetzung zuerst das Drift Policy Gate ausgeben.
- Bei
feat:niemals Evidence-Datei manuell erstellen; immerscripts/generate_feature_evidence.pyverwenden. - Bei
feat:oderfix:immerCHANGELOG.mdaktualisieren. - Push-Gates sind bindend; Bypass nur als Notfall und dokumentiert.
- Fuer unklare Zuordnung immer auf
.github/instructions/drift-context-routing.instructions.mdwechseln.
Pflicht-Shortcuts: Wann welchen make-Befehl aufrufen
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.
- 3d ago First seen · 92 lines · 1,332 tokens per session scan A b427a4b85f5c
drift drift-agent-quickref.instructions.md is an instructions file published in the GitHub repository mick-gsk/drift (14 stars, last pushed 1mo ago), licensed MIT. It adds 1,332 tokens to every session, about $0.0067 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.
Other instructions, from other repositories
truecourse CLAUDE.md
Instructions for truecourse-ai/truecourse, covering truecourse — claude instructions, key files to keep updated, project layout, development commands and storage.
truecourse AGENTS.md
Instructions for truecourse-ai/truecourse, covering truecourse — codex instructions, key files to keep updated, project layout, development commands and rules.
PhpCodeArcheology CLAUDE.md
Instructions for PhpCodeArcheology/PhpCodeArcheology, covering claude.md — phpcodearcheology, project overview, language, tech stack and project structure.
vibe-check AGENTS.md
Instructions for benavlabs/vibe-check, covering security rules, secrets, database, authentication and authorization and input and output.
Specflow CLAUDE.md
Instructions for Hulupeep/Specflow, covering claude.md, project context, specflow rules, rule 1: no ticket = no code and rule 2: commits must reference an issue.
arai CLAUDE.md
Instructions for taniwhaai/arai, covering claude.md — arai, commands, architecture, prompt-collector module (src/promptcollector.rs) and extending the match pipeline.