Skip to content

docs(readme): add Getting Started guide with fork, layout, and env var setup#3

Open
sixty4bit wants to merge 6 commits intoworkos:mainfrom
sixty4bit:docs/improve-readme-setup
Open

docs(readme): add Getting Started guide with fork, layout, and env var setup#3
sixty4bit wants to merge 6 commits intoworkos:mainfrom
sixty4bit:docs/improve-readme-setup

Conversation

@sixty4bit
Copy link

Summary

  • Replaced the bare "Quick Start" section with a structured "Getting Started" guide covering prerequisites that were previously undocumented
  • Added step-by-step instructions for: forking the repo, sibling directory layout, environment variable setup (CASE_ASSETS_REPO, CASE_LEARNINGS_REPO), plugin installation
  • Added read-learning.sh and write-learning.sh to the file structure listing
  • Updated the self-improvement section to reflect that learnings are stored in an external repo

Why

New users had no guidance on forking (required for the self-improvement loop), the sibling directory layout (required for projects.json paths to resolve), or the env vars (required for screenshot uploads and learnings). These gaps were discovered while onboarding to the project.

Test plan

  • Verify markdown renders correctly on GitHub (headings, table, code blocks, anchor links)
  • Follow the Getting Started steps on a fresh machine to confirm they work end-to-end

🤖 Generated with Claude Code

sixty4bit and others added 6 commits March 17, 2026 10:40
All references to /Users/nicknisi/Developer/case have been replaced
with dynamic resolution so the harness works for any user/fork.

Shell scripts and hooks now derive CASE_REPO from their own location
using the SCRIPT_DIR pattern (matching bootstrap.sh). SKILL.md and
agent prompts use ${CASE_REPO} as a variable resolved by the
orchestrator at invocation time. Each agent's Input contract now
includes CASE_REPO as an explicit parameter.

upload-screenshot.sh now requires the CASE_ASSETS_REPO env var
instead of defaulting to a specific user's repo, with clear setup
instructions on failure.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
fix(harness): remove hardcoded filesystem paths for portability
…b repo

Per-repo tactical knowledge no longer lives in docs/learnings/{repo}.md
inside the case repo. Instead, learnings are stored in an external GitHub
repo configured via the CASE_LEARNINGS_REPO env var, matching the
CASE_ASSETS_REPO pattern.

New scripts:
- scripts/read-learning.sh: reads learnings via gh api (stdout)
- scripts/write-learning.sh: appends entries and commits via gh api

The implementer agent now calls read-learning.sh before coding, and the
retrospective agent calls write-learning.sh to persist tactical knowledge.
Harness-level escalations (3+ similar learnings → convention) still write
to the local case repo.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
refactor(learnings): externalize per-repo learnings to separate GitHub repo
…r setup

The Quick Start section assumed users already had a fork, knew about
the sibling directory layout, and had env vars configured. New users
had no guidance on any of these prerequisites.

Replaced Quick Start with a structured Getting Started that covers:
1. Fork the repo (and why — self-improvement loop)
2. Clone target repos alongside case (sibling layout diagram)
3. Set up environment variables (CASE_ASSETS_REPO, CASE_LEARNINGS_REPO)
4. Install the plugin
5. Use with an issue

Also added read-learning.sh and write-learning.sh to the file
structure listing, and updated the self-improvement section to
reflect that learnings are stored in an external repo.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
projects.json contains user-specific paths that vary per checkout.
Without gitignoring it, every fork would push their local paths
upstream.

Now projects.json follows the .env/.env.example pattern:
- projects.example.json ships with the repo (default sibling layout)
- projects.json is gitignored (user copies and customizes paths)

Updated Getting Started, CLAUDE.md, and file structure listings
to reflect the new pattern.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
nicknisi added a commit that referenced this pull request Mar 17, 2026
- Remove double-written metrics: drop legacy log-run.sh call from
  pipeline.ts, writer.ts is the single path (#1)
- Await retrospective instead of fire-and-forget so process.exit
  doesn't kill it (#3)
- Fix README claiming retrospective applies changes directly — it
  proposes amendments, only learnings are applied directly (#6)
- Document attended-mode retry semantics: maxRetries is per-attempt,
  human can re-enter implement indefinitely (#2)
- Reconcile --worktree docs: removed from flag list, branch isolation
  is handled by skill layer before orchestrator dispatch (#4)
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Development

Successfully merging this pull request may close these issues.

1 participant