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/sidcorp-co/forge/claude-mdgit clone --depth 1 https://github.com/SidCorp-co/forgeWhat 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 | $0.02344 | $0.02344 |
| Opus 5 | $0.01172 | $0.01172 |
| Sonnet 5 | $0.00469 | $0.00469 |
| Haiku 4.5 | $0.00234 | $0.00234 |
Grade A, and why
forge CLAUDE.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 2d 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 — 147 lines — stays where its author put it; the contents beside it link to each section on GitHub.
@.forge/orientation.md
Forge
Open-source control plane for Claude Code — full-stack project management + an AI agent pipeline that drives Claude end-to-end.
Constitution: docs/VISION.md — what Forge is / is not, why, who, and the
principles (incl. VISION: state-never-lies, VISION: kernel-hard-policy-soft). Intent only: no architecture,
no versions, no roadmap. On intent conflicts, VISION wins — cite it by name, never by section
number.
Workspace
| Package | What |
|---|---|
packages/core |
Hono backend. Single app (src/index.ts) mounting per-domain route modules (src/<domain>/routes.ts); Drizzle ORM over Postgres (pgvector); WebSocket server (/ws); MCP server (/mcp, tools in src/mcp/tools/forge-*.ts); the pipeline dispatcher that drives Claude. |
packages/web-v2 |
Next.js cloud UI, canonical at /. Feature modules under src/features/<domain>/. |
packages/runner |
Headless Rust forge-runner CLI daemon (crates forge-runner / forge-runner-core) for servers/CI; pairs as a device. |
packages/contracts |
Shared cross-app TS types & registries (issues.ts, pipeline-registry.ts, requests.ts, responses.ts, rows.ts, domain-templates.ts). |
packages/observability |
Shared telemetry helpers (incl. the secret scrubber). |
Commands
pnpm verify when you finish coding, before you push — the conformance entrypoint. It runs
every check CI runs, reports all of them in one pass instead of stopping at the first, and prints
the cm:guard / cm:edge / cm:flow declared on the files you touched. Exit 0 clean · 1
violations · 2 a check could not run. Hooks only make it arrive sooner; a contributor with no
plugin installed is held to exactly the same bar.
From the repo root, turbo fans out: pnpm dev / pnpm build / pnpm test / pnpm typecheck /
pnpm lint. Per package (from inside packages/<pkg>/):
| Package | Dev | Build | Test | Lint |
|---|---|---|---|---|
| core | pnpm dev (tsx watch) |
pnpm build (tsc) |
pnpm test (vitest); pnpm test:integration |
pnpm lint (biome) |
| web-v2 | pnpm dev (next, :3100) |
pnpm build |
pnpm test (vitest) |
pnpm lint (check-lint-budget, not biome directly) |
| runner | — | cargo build (in packages/runner) |
cargo test |
— |
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.
- 2d ago First seen · 147 lines · 2,344 tokens per session scan A 2e0a249495c2
forge CLAUDE.md is an instructions file published in the GitHub repository SidCorp-co/forge (5 stars, last pushed 3d ago), licensed Apache-2.0. It adds 2,344 tokens to every session, about $0.0117 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 2026-08-31.
Other instructions, from other repositories
rondoflow CLAUDE.md
Instructions for rondoflow/rondoflow, covering rondoflow, project overview, tech stack, development and setup.
OpenVitamin AGENTS.md
Instructions for fengzhizi715/OpenVitamin, covering agents.md — inference gateway & plugin governance, 1. 项目与技术栈对齐说明, 2. 核心设计原则（core principles）, 2.1 user-in-control（用户始终掌控） and 2.2 gateway-centric（以推理网关为中心）.
AIWorkHub AGENTS.md
Instructions for shrec/AIWorkHub, covering aiworkhub mcp agent notes and aiworkhub mcp tool-use policy.
orcho-mcp AGENTS.md
Instructions for symphos-ai/orcho-mcp, covering orcho-mcp instructions, scope, local instruction files, workspace development pipeline and stable install is read-only.
AIWorkHub copilot-instructions.md
Instructions for shrec/AIWorkHub, a project described as: Open-source, local-first control plane for multi-model AI coding agents in VS Code and MCP: task DAGs, source intelligence, durable context and evidence-based review.
sync AGENTS.md
AGENTS.md instructions for AgriciDaniel/sync, covering sync agent instructions, read order, operating rules and verification.