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/icoretech/airbroke/agents-mdgit clone --depth 1 https://github.com/icoretech/airbrokeWhat 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 | $0.01096 | $0.01096 |
| Opus 5 | $0.00548 | $0.00548 |
| Sonnet 5 | $0.00219 | $0.00219 |
| Haiku 4.5 | $0.00110 | $0.00110 |
Grade A, and why
airbroke 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.
How it starts
The opening of the file, as written. The whole thing — 98 lines — stays where its author put it; the contents beside it link to each section on GitHub.
PROJECT KNOWLEDGE BASE
Generated: 2026-07-24 Commit: 8a95b54b Branch: main
OVERVIEW
Airbroke is a self-hosted error collector and triage UI built with Next.js 16, React 19, Prisma 7, and PostgreSQL 18. Development and verification are Docker Compose-first; the repository pins Yarn 4.17.1.
STRUCTURE
app/ # App Router pages and external HTTP protocol handlers
components/ # Maintained feature UI plus registry-managed primitives
lib/ # Auth, actions, queries, intake, MCP, and shared server logic
prisma/ # Schema, forward migrations, seed, and generated client output
__tests__/ # Central Vitest suite mirroring runtime boundaries
docs/agents/ # Detailed cross-cutting guidance linked below
WHERE TO LOOK
| Task | Location | Notes |
|---|---|---|
| Routes and layouts | app/ |
Read app/AGENTS.md |
| Intake | app/api/, lib/intake/ |
Synchronous writes |
| Auth | proxy.ts, lib/auth/ |
Server checks are authoritative |
| UI | components/ |
Read components/AGENTS.md |
| Mutations | lib/actions/ |
Auth plus explicit invalidation |
| UI reads | lib/queries/ |
MCP has separate read models |
| MCP API | app/api/mcp/, lib/mcp/ |
Read-only tools and static-key auth |
| Data model | prisma/ |
Read prisma/AGENTS.md |
| Tests | __tests__/ |
Read __tests__/AGENTS.md |
CODE MAP
Reference counts are unmeasured; relationships below come from direct source imports and route delegation.
| Symbol | Type | Location | Refs | Role |
|---|---|---|---|---|
db |
Prisma proxy | lib/db.ts |
unmeasured | Lazy shared database client |
getAuth |
Factory | lib/auth/index.ts |
unmeasured | Lazy auth setup |
requireAuth |
Guard | lib/actions/requireAuth.ts |
n/a | Action auth |
upsertNoticeOccurrence |
Service | lib/intake/ |
n/a | Intake writes |
MCP_TOOLS |
Registry | lib/mcp/tools.ts |
unmeasured | MCP contract |
handleMcpPost |
Handler | lib/mcp/server.ts |
unmeasured | MCP transport |
revalidateProjectShellPaths |
Helper | lib/actions/ |
n/a | Invalidation |
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 · 98 lines · 1,096 tokens per session scan A 092a2883fc39
airbroke AGENTS.md is an instructions file published in the GitHub repository icoretech/airbroke (222 stars, last pushed 3d ago), licensed MIT. It adds 1,096 tokens to every session, about $0.0055 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
LaunchStack CLAUDE.md
Instructions for Deodat-Lawson/LaunchStack, covering claude.md, frontend rules (apps/web, apps/landing) and workflow.
tutor-management-app CLAUDE.md
Instructions for t1112000/tutor-management-app, covering claude.md, commands, architecture, auth and page pattern.
tutor-management-app AGENTS.md
Instructions for t1112000/tutor-management-app, covering repository guidelines, project structure & module organization, build, test, and development commands, coding style & naming conventions and testing guidelines.
agent-stack-template AGENTS.md
Instructions for krivoox/agent-stack-template, covering this is not the next.js you know, agent guide, what this repository is, every turn and what to open after classify.
MagicAIBuilder AGENTS.md
Instructions for KaelSensei/MagicAIBuilder, covering magicaibuilder — coding standards, language & runtime, type safety — non-negotiable rules, no type assertions (as) and no non-null assertions (!).
openmonetis AGENTS.md
AGENTS.md instructions for felipegcoutinho/openmonetis, covering agents.md - openmonetis, related projects, critical rules, architecture and feature-first.