docs: full rewrite around one brain-first story, with professional diagrams#38
Merged
CodeWithJuber merged 1 commit intoJul 8, 2026
Merged
Conversation
…agrams
Rewrites the primary + community docs so a newcomer immediately understands
what forgekit is and does. Single positioning through-line ("one brain for
every AI coding agent" — the cognitive substrate a stateless model lacks:
memory, foresight, guardrails; cross-tool config is the delivery mechanism),
one voice, one visual system.
- New showpiece hero (docs/assets/hero-{dark,light}.svg), embedded at the
README top via the GitHub prefers-color-scheme <picture> pattern, on the
live forge-dash palette so docs and product read as one system.
- All six mermaid diagrams restyled with one shared brand theme block +
a single ember accent node, so every diagram looks identical. ARCHITECTURE
gains three new accurate diagrams: the pre-action gate pipeline, the PCM
ledger + team merge, and the reuse/context loop.
- README: brain-first hook, TOC + H3s for skimming, day-to-day value before
the benchmark numbers, commands table regrouped and verified against
src/cli.js (every documented verb is real).
- ARCHITECTURE: rebuilt around the four flows; fixes accuracy drift — the
code graph is .forge/atlas.json (not SQLite/.db), runtime is
zero-dependency (serena/LSP moved to an honest "not shipped" note).
- ONBOARDING/GUIDE: brain-first intros, command-group indexes, restyled
diagrams, one-liners verified against source.
- ACCESSIBILITY (stub -> honest statement of what forge uicheck actually
gates), examples/README (stub -> two real worked examples), ROADMAP /
CONTRIBUTING / SUPPORT / RELEASING voice + accuracy pass (Node 20/22).
- One canonical tagline aligned across brand.json, package.json, CITATION.cff
(and it already matches the landing page).
Docs-only; 438 tests still pass, typecheck and biome clean.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01Fc2MvWJbQ1cyNQ76ymv4hs
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
What & why
The docs were factually correct (after #37) but hard to grok: five conflicting taglines split between "one config" and "one brain," a dense flat README, and plain default-mermaid diagrams. This is a full rewrite of the primary + community docs so a newcomer immediately understands what forgekit is and does — one positioning, one voice, one visual system.
The single through-line (brain-first): a large language model is stateless — one context window, wiped every call, with no memory, no foresight, no enforced guardrails. forgekit is the cognitive substrate that supplies exactly those three things; the cross-tool config compiler is how that brain gets delivered into 9 tools at once. "One config for every tool" is now the delivery mechanism, not the headline.
Highlights
docs/assets/hero-dark.svg+hero-light.svg), embedded at the README top via the GitHubprefers-color-scheme<picture>pattern, drawn on the liveforge dashpalette so docs and product read as one system. Rendered and eyeballed in both light and dark.src/cli.js— every documentedforge <verb>is a real command..forge/atlas.json(not SQLite/.db), runtime is zero-dependency (serena/LSP moved to an honest "not shipped, exploring" note).forge uicheckactually gates),examples/README(stub → two real copy-pasteable examples), ROADMAP / CONTRIBUTING / SUPPORT / RELEASING voice + accuracy pass (Node 20/22, not 18/20/22).brand.json,package.json,CITATION.cff(and it already matches the landing page).Checklist
npm testpasses — 438 pass / 0 fail / 2 skip (unchanged; the 2 skips are the Playwright-optional tier)npm run checkpasses (Biome) — clean on the touched JSON (package.json,brand.json)docs:)CHANGELOG.mdupdated under## [Unreleased]— n/a, docs clarity pass; no shipped behavior changenpm run typecheckpassesRisk & rollback
Extra checks
src/cli.js'sCOMMANDS; no surviving.forge/atlas.db/serena/LSP-as-shipped claims; all relative links and the newdocs/assets/hero-*.svgpaths resolve; both SVGs are well-formed and render in light + dark.Generated by Claude Code