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/thinkinaixyz/deepchat/agents-mdgit clone --depth 1 https://github.com/ThinkInAIXYZ/deepchatWhat 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 | $0.00443 | $0.00443 |
| Opus 5 | $0.00221 | $0.00221 |
| Sonnet 5 | $0.00089 | $0.00089 |
| Haiku 4.5 | $0.00044 | $0.00044 |
Grade A, and why
deepchat AGENTS.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.
What it actually says
Repository Guidelines
- Prefer the smallest correct change; add no abstraction or dependency without a real need.
- Preserve unrelated worktree changes; avoid destructive Git unless explicitly requested.
- Use pnpm only; require Node >=20.19 and pnpm >=10.11.
- Core code:
src/main,src/preload,src/renderer,src/shared; stack: Electron/Vue 3/TS. - Tests: use Vitest in
test/main, Vitest with Vue Test Utils intest/renderer, and Playwright intest/e2e; run the smallest relevant suite. - Default to implementation-first development: inspect and design, complete the planned code change, then decide what new verification is needed. Do not default to TDD unless explicitly requested or a minimal executable reproduction is required to understand a complex failure.
- Existing checks may run at any time. Treat new tests as regression protection, not implementation scaffolding; remove temporary probes and tests before handoff.
- Commit only the smallest durable tests for user-visible behavior, documented contracts, persistence or migration, lifecycle or concurrency, recovery, security boundaries, or proven regressions. Prefer no new test to implementation-coupled coverage.
- Keep native capabilities behind typed preload/IPC boundaries with context isolation enabled.
- Use vue-i18n for user copy; prefer existing shadcn-vue primitives and VueUse utilities.
- Follow Oxfmt: single quotes, no semicolons, 100 columns.
- Before editing, inspect ownership, callers, tests, and nearby patterns; fix the root cause.
- Before handoff run format, i18n, lint, typecheck, and relevant test suites.
- Keep provider and ACP registry refreshes produced by normal builds.
- Use Conventional Commits (
type(scope): subject, <=50 chars); never add AI co-authors. - Target routine PRs to
dev; only release branches targetmain. - Assess scope first; use SDD only for large features or complex changes, never small local work.
- For UI changes, include concise BEFORE/AFTER ASCII layouts in the PR.
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 · 27 lines · 443 tokens per session scan A 80c4bfe06348
deepchat AGENTS.md is an instructions file published in the GitHub repository ThinkInAIXYZ/deepchat (6,302 stars, last pushed yesterday), licensed Apache-2.0. It adds 443 tokens to every session, about $0.0022 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
Librechat-Mobile CLAUDE.md
Instructions for garfiec/Librechat-Mobile, covering switchboard, tech stack, module layout, architecture rules and adding a new feature module.
ExtremeRouter AGENTS.md
Instructions for rsalmn/ExtremeRouter, covering extremerouter — lead engineer agent, core philosophy, thinking process, critical thinking and coding standards.
jobs-mcp-proxy CLAUDE.md
Claude Code instructions for foundrole/jobs-mcp-proxy, covering claude.md - jobs mcp proxy project development rules, project overview, package manager, ✅ correct and ❌ incorrect.
rigorous-agents-md AGENTS.md
Instructions for Zkeeper6899/rigorous-agents-md, covering rigorous agents.md, rigorous reasoning and engineering judgment, communication style and optional domain extensions.
purmemo-mcp CLAUDE.md
Claude Code instructions for purmemo-ai/purmemo-mcp, covering purmemo project instructions, session startup: get project context, context gathering (start of every task), option 1: use the research workflow (recommended) and option 2: manual research (if skill not available).
langgraph AGENTS.md
AGENTS.md instructions for langchain-ai/langgraph, covering agents instructions, corridor security analysis, libraries and dependency map.