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 commands/dormstern/forge/helpgit clone --depth 1 https://github.com/dormstern/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 | $0.00055 | $0.00912 |
| Opus 5 | $0.00028 | $0.00456 |
| Sonnet 5 | $0.00011 | $0.00182 |
| Haiku 4.5 | $0.00006 | $0.00091 |
Grade A, and why
help 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 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.
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
/forge:help — Orientation
Execution mode: direct. This command has explicit steps below — execute them in order. Do not enter plan mode, do not write a plan file, do not call ExitPlanMode. The framework already did the planning; you just run the steps.
You are introducing Forge to a user who has the plugin installed but isn't sure how to use it. Print the orientation block below, exactly as written. Do NOT add commentary, do NOT modify the structure, do NOT skip sections. Use the user's actual project context (existence of IDENTITY.md / THESIS.md) to set the Where you are now section, but everything else is verbatim.
═══════════════════════════════════════════════════════
FORGE — Vibe-coded plans go in. Only what survives ships.
═══════════════════════════════════════════════════════
THREE FILTERS EVERY RELEASE RUNS THROUGH:
1. Question every plan — Devil's Advocate before code exists
2. Delete 10% of scope — mandatory, every release
3. Audit every five features — drift checks vs IDENTITY
Then the Ralph Loop ships what survived.
TWO MODES (auto-routed by which state file exists in your project):
┌──────────────────────┬──────────────────────────────────┐
│ FOUNDER mode │ PM mode │
│ Pre-PMF validation │ Shipping a product │
│ THESIS.md present │ IDENTITY.md present │
│ │ │
│ /forge:scout │ /forge:harvest │
│ /forge:interrogate │ /forge:shape ← hero command │
│ /forge:synthesize │ /forge:build │
│ /forge:graduate │ /forge:ship │
└──────────────────────┴──────────────────────────────────┘
WORKS IN BOTH MODES:
★ /forge:try 30-second magic-moment demo. Zero setup. Start here.
/forge:falsify Devil's Advocate against any plan / pitch / PRD
/forge:init Scaffold state files (Validating by default)
/forge:init --pm Scaffold for shipping a product instead
/forge:cycle Run a full cycle, auto-routed
/forge:drift Standalone drift audit on last 5 features
/forge:status One-screen state snapshot
/forge:help This screen
STATE FILES LIVE IN YOUR REPO. Plain Markdown + JSON:
PM mode (6 files):
IDENTITY.md, features.json, releases.json,
progress.md, architecture.md, contracts.md
Founder mode (7 files):
THESIS.md, hypotheses.json, evidence.json,
cycles.json, interviews.md, landscape.md, progress.md
Commit them or .gitignore them — they're yours. Claude reads
them on every session via the SessionStart hook.
═══════════════════════════════════════════════════════
After printing the block, append a single contextual Where you are now: section:
- If
IDENTITY.mdexists in the project root: "PM mode detected. Run/forge:statusto see your current cycle, or/forge:harvestto start a new one." - If
THESIS.mdexists: "Founder mode detected. Run/forge:statusto see hypothesis trajectory, or/forge:interrogateto keep validating." - If both exist: "Both modes detected. Tell me which to run — or rename one of the state files to disable that mode."
- If neither exists: "No state files yet. Run
/forge:init(PM, default) or/forge:init --founder(pre-PMF) to scaffold."
End with one line: "Read more: github.com/dormstern/forge"
Do NOT execute any other command. Do NOT make recommendations beyond the contextual one-liner above.
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 · 77 lines · 55 tokens per session scan A 282931a1e370
help is a command published in the GitHub repository dormstern/forge (6 stars, last pushed 3mo ago), licensed MIT. It adds 55 tokens to every session and 912 once invoked, about $0.0003 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 commands, from other repositories
checklist
Generate a custom checklist for the current feature based on user requirements.
clarify
Identify underspecified areas in the current feature spec by asking up to 5 highly targeted clarification questions and encoding answers back into the spec.
specify
Create or update the feature specification from a natural language feature description.
analyze
Perform a non-destructive cross-artifact consistency and quality analysis across spec.md, plan.md, and tasks.md after task generation.
converge
Assess the current codebase against the feature's spec, plan, and tasks, then append any remaining unbuilt work as new tasks to tasks.md so implement can complete it.
implement
Execute the implementation plan by processing and executing all tasks defined in tasks.md.