ContextForgeShip coding-agent context from your real tree — not a guess.
Paste a file tree (and optional key files). Get AGENTS.md, rules files, and a machine-readable spec as a reviewable draft — never a silent mock.
No credit card required · Cancel anytime
Key Takeaways
- Deterministic tree extraction before any model prose.
- Outputs the instruction files agents actually read (AGENTS.md, Cursor, Claude, Copilot).
- Draft packs for review — not a guarantee agents stop hallucinating.
Your tree. The instruction files agents expect.
Decision-support drafts — not a certificate that agents behave.
ContextForge — definition
ContextForge is a generator that converts a codebase snapshot — file tree plus optional key files — into the context files coding agents read: AGENTS.md, .cursorrules, CLAUDE.md, .github/copilot-instructions.md, and a machine-readable spec.
What this is NOT
- Not a guarantee that coding agents stop hallucinating.
- Not a silent commit. Review every inferred convention before you commit the files.
- Not a guarantee agents stop hallucinating — review drafts before production use.
Controller / Processor — When packs mention privacy roles, we use GDPR Art.4 labels: Controller and Processor (decision-support only).
Agents read files. Guesswork leaves gaps.
Missing instruction files
Repos ship without AGENTS.md, Cursor rules, or Copilot instructions.
Stale or invented paths
Free-form LLM guesses invent files that are not in the tree.
Silent mock packs
Fake success responses hide that AI was never configured.
No review trail
Teams cannot tell what was extracted vs invented.
Paste tree. Draft pack. Review.
Deterministic extraction first; labelled model prose second.
Tree-first extract
Paths and structure come from what you paste — not invention.
Agent instruction files
AGENTS.md, rules, Copilot instructions, and a reviewable spec.
Honest demo vs live
Demo is labelled; live narrative needs a configured key.
Exportable draft
A pack you can edit and commit — not a silent guess.
Everything you need to draft AI-ready context
Built for teams standardising Cursor/Claude/Copilot context across repos.
ContextForge
Paths and structure come from what you paste — not invention.
ContextForge
AGENTS.md, rules, Copilot instructions, and a reviewable spec.
ContextForge
Demo is labelled; live narrative needs a configured key.
ContextForge
A pack you can edit and commit — not a silent guess.
From tree to draft pack in 3 steps
Paste
Paste a file tree and optional key file contents.
Extract
Deterministic structure plus optional model-assisted prose.
Review
Edit and commit — humans stay in the loop.
Paste tree. Draft pack. Review.
Deterministic extraction first; labelled model prose second.
Missing instruction files
Repos ship without AGENTS.md, Cursor rules, or Copilot instructions.
Tree-first extract — Paths and structure come from what you paste — not invention.
Try it on this page
or generate below.
Saved history
What the studio actually drafts
Capability cards, not customer quotes. You still review inferred conventions before commit.
“Paths, package manager, framework, tests, and CI come from the tree you paste.”
“Drafts AGENTS.md, .cursorrules, CLAUDE.md, copilot-instructions.md, and GEMINI.md.”
“Detected facts and inferred guesses are marked. You review before commit.”
Honest boundaries
ContextForge drafts instruction files from your tree. Customize and review — no efficacy guarantees.
Simple plans that scale
All paid plans include
- Context pack studio
- Deterministic tree extract
- Labelled model-assisted prose (decision-support)
- Email support on paid plans — no hidden add-ons for core studio use
- AI drafts 3/day · 15/mo
- Studio demo (no extra LLM fee)
- AI drafts 40/day · 300/mo
- No separate ChatGPT subscription required
- Priority support · 2 months free with yearly
Feature comparison
What each plan includes — quotas reflect the enforced fair-use limits.
| Dimension | Free | Pro | Enterprise |
|---|---|---|---|
| Usage | |||
| AI runs | 3/day · 15/mo | 40/day · 300/mo | 200/day · 3,000/mo |
| Report export | Watermarked | Clean export | Clean export |
| Audit log | Not included | ✓ Included | ✓ Included |
| Product | |||
| Core studio workflow | ✓ Included | ✓ Included | ✓ Included |
| Deterministic checklist | ✓ Included | ✓ Included | ✓ Included |
| Exportable report | ✓ Included | ✓ Included | ✓ Included |
| Decision-support framing | ✓ Included | ✓ Included | ✓ Included |
| BYOK (your LLM key) | Not included | Not included | ✓ |
| SSO | Not included | Not included | Roadmap |
| Support & compliance | |||
| Support | Community | Priority | Custom SLA |
Already paid? Unlock Pro with your license
Licenses are issued manually after your payment is confirmed: we generate one per paid order and send it to the purchase email, typically within a few hours. Enter it below to activate Pro on this device — no account needed. Nothing yet? Write to lixingliangsy@163.com and reference your order email.
Your questions, answered
What does ContextForge generate?
A draft context pack (AGENTS.md, rules, Copilot instructions, spec) from your tree — reviewable, not a silent guess or efficacy guarantee.
Is this legal advice or a certification?
No. Outputs are decision-support drafts. Have counsel and risk owners review before production. We do not claim ISO, CMP, or regulator certification.
Can I cancel anytime?
Yes. Self-serve plans cancel anytime; access continues until period end.
Do I need a credit card to start?
No. Start with email signup or Demo mode in the studio, then upgrade when ready.
Do I need my own ChatGPT / OpenAI subscription?
No for Free/Pro. AI runs are included in your plan (fair use) via our platform key. Enterprise can optionally bring your own OpenAI key (BYOK) — configure it at /settings (key stays server-side only).
What is Fair Use / what if I hit the AI limit?
Pro includes 40 AI drafts per day and 300 per month. If you hit the fair-use limit, Studio returns a labelled rule-based draft until reset — or upgrade / use Enterprise BYOK for heavier volume.
Is checkout secure?
Payments are processed by Waffo Pancake (merchant of record).
What happens after I pay?
You receive access confirmation; fulfillment is tracked via webhook + order logs.
Quick answers
- Paste your file tree (git ls-files / find output)
- Optionally paste 1-3 key files (package.json, entry point, config)
- Pick the target agents (Cursor, Claude Code, Codex, Copilot, Gemini)
- ContextForge extracts stack facts deterministically, then drafts the files
- Review, edit, commit — the output is a draft, not an oracle
More questions
What inputs?
A file tree and optional key file contents you paste — no repo credentials required for the draft path.
Does it guarantee better agents?
No. It drafts instruction files; humans review before use. No efficacy claims.
Deterministic vs model?
Structure and paths are rule-based; prose sections may be model-assisted and labelled.
Demo vs live?
Demo mode is labelled; live narrative needs a configured AI key.
Who is it for?
Teams standardising Cursor/Claude/Copilot context across repos.
Primary sources?
Agent instruction file conventions (AGENTS.md, .cursorrules, CLAUDE.md, Copilot).
How it compares
| Dimension | Manual | ContextForge |
|---|---|---|
| Speed | Hours to days | Minutes per run |
| Consistency | Varies by person | Same ruleset every run |
| Output | Free-form | Structured, exportable result |
| Best for | Final sign-off | First-pass decision-support |
When not to use: Do not commit the generated files unread. A context file that states conventions your repo does not follow is worse than no context file. ContextForge labels every inferred fact as inferred so you can correct it.
- deterministic extraction vs model guess 2026 — From the ContextForge blog.
- what an ai ready context pack contains 2026 — From the ContextForge blog.
- Fleet launch notes — ContextForge ships in the LX AI Micro-SaaS fleet.
People also search
Click a tab to expand it; click again to hide. All content stays on the page for search engines.
Start smarter today
Join builders using ContextForge. Free to try — no card needed.
Trusted · Cancel anytime