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/dmno-dev/varlock/copilot-instructionsgit clone --depth 1 https://github.com/dmno-dev/varlockWhat 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 | $0.00474 | $0.00474 |
| Opus 5 | $0.00237 | $0.00237 |
| Sonnet 5 | $0.00095 | $0.00095 |
| Haiku 4.5 | $0.00047 | $0.00047 |
Grade A, and why
varlock copilot-instructions.md scanned grade A 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 yesterday.
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.
Runs shell commandslowCapability
Expected in a hook, worth knowing in a rule or an instructions file.
6. **Risky APIs:** Flag unsafe `child_process`, `vm`, or deserialization from untrusted data. How it starts
The opening of the file, as written. The whole thing — 36 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Copilot PR code review (security)
Scope: Focus on code security. Copilot does not review dependency manifests or lockfiles; rely on Dependabot, CodeQL, and the dependency-review workflow for supply chain.
Security checklist (changed lines)
- Secrets: No hardcoded API keys, tokens, passwords, or private URLs. Preserve redaction and leak prevention for sensitive env values.
- Injection / parsing: Safe handling of user-controlled strings in @env-spec parsing and loaders; avoid
eval/new Functionon untrusted input. - Process execution:
exec/spawn/ shell (secret plugins, integrations): avoid shell injection; validate arguments; no untrusted paths. - Filesystem: No path traversal or arbitrary read/write from external input.
- Trust boundaries: Document what untrusted config can do; avoid unsafe defaults.
- Risky APIs: Flag unsafe
child_process,vm, or deserialization from untrusted data. - Crypto: Prefer standard libraries; avoid ad-hoc cryptography.
- AI safety: No prompt injection or other unsafe patterns that might allow exfiltration of secrets.
Varlock: Loaders, CLI, packages/plugins, and MCP integrations are high impact.
Copilot Instructions for Varlock
Varlock adds declarative schema to .env files using @env-spec decorator comments — validation, coercion, type safety, redaction/leak prevention for sensitive values, and multi-environment management.
All project conventions live in AGENTS.md — repo structure, bun/workspace usage, scripts, testing, versioning (bumpy changesets), linting, and branch/PR rules. Follow it.
Quick reference:
bun install # install dependencies
bun run build:libs # build all libraries (excludes website)
bun run test:ci # run tests once (CI mode)
bun run lint:fix # lint with auto-fix — run before completing any task
Before finishing any task: run bun run lint:fix, verify bun run build:libs passes, and add a bumpy changeset if publishable packages changed (see AGENTS.md for details).
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.
- yesterday First seen · 36 lines · 474 tokens per session scan A 10334e17e52e
varlock copilot-instructions.md is an instructions file published in the GitHub repository dmno-dev/varlock (4,238 stars, last pushed 2d ago), licensed MIT. It adds 474 tokens to every session, about $0.0024 per session on Opus 5. A static security scan graded it A with 1 finding (runs shell commands). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other instructions, from other repositories
tweakcc AGENTS.md
Instructions for Piebald-AI/tweakcc, covering agents.md, development commands, build, testing and linting & formatting.
code-abyss CLAUDE.md
Instructions for telagod/code-abyss, covering claude.md, project overview, commands, persona behavioral battery (opt-in eval) and ci / cd.
envault AGENTS.md
Instructions for Coding-Dev-Tools/envault, covering envault — agents.md, quick start, commands, development and install dev deps.
configdrift AGENTS.md
Instructions for Coding-Dev-Tools/configdrift, covering configdrift, purpose, build & test commands, architecture and conventions.
claude-calibration CLAUDE.md
Claude Code instructions for odere-pro/claude-calibration, covering claude-calibration — plugin development, what ships, what doesn't, pointers and agent routing.
cmds-llm-wiki AGENTS.md
Schema and harness document for the CMDS LLM Wiki vault. Defines the 3-layer architecture (Raw Sources / Wiki / Schema), ingest-query-lint operations, file conventions, and frontmatter standards. This is the single source of truth for LLM behavior in this vault.