Skip to content

Engine-only install (4) README and ecosystem text for engine-only mode - #467

Closed
EdbertChan wants to merge 4 commits into
plan/engine-only-install-3-prune-dangling-repo-owned-symlinks-and-dead-hook-entriesfrom
plan/engine-only-install-4-readme-and-ecosystem-text-for-engine-only-mode
Closed

Engine-only install (4) README and ecosystem text for engine-only mode#467
EdbertChan wants to merge 4 commits into
plan/engine-only-install-3-prune-dangling-repo-owned-symlinks-and-dead-hook-entriesfrom
plan/engine-only-install-4-readme-and-ecosystem-text-for-engine-only-mode

Conversation

@EdbertChan

@EdbertChan EdbertChan commented Sep 12, 2026

Copy link
Copy Markdown
Owner

Summary

Describe ./install.sh --engine-only in README.md and docs/ecosystem.md so a
reader knows what the bare engine links, what it prunes, and that corpus
stays in git and refills as reflect and automate-me run. Depends on the
dead-hook prune slice having merged so the text matches shipped behavior.

Goal: A reader can find the engine-only mode from the README install
section and understand it from the ecosystem overview.
Motivation: The flag exists but nothing in the repo text mentions it.
Safety invariant: Text only; install.sh, tests, hooks, and skills are
untouched.
Effectiveness measurement: grep -c -- '--engine-only' README.md docs/ecosystem.md goes from 0 in each file to at least 1 in each.

Test Plan

Test Plan
  • grep -q -- '--engine-only' README.md && grep -q -- '--engine-only' docs/ecosystem.md — Review claim: Both files mention the flag and nothing else changed.
    Review lane: proof
    Safety invariant: Read-only check.
    Effectiveness measurement: The grep counts are the measurement.
    Slice rationale: One focused proof for the text slice.
    Architectural effect: None; verification only.
    Goal: Prove both files mention the flag.
    Motivation: A text change is only reviewable when its presence is machine-checked.
    Alternative considerations: Manual reading was ruled out as non-deterministic.
    Implementation details: Run the two greps.
    Non-goals: No mutations.
    Layer: docs
    Feature state: active
    Acceptance criteria:

  • Exits 0 only when both files mention the flag.

  • bash scripts/scrub-handoff-artifacts.sh — Review claim: No ephemeral inter-task handoff files remain in the worktree before the merge gate.
    Review lane: cleanup
    Safety invariant: The scrub script only removes known handoff artifact names and never touches source, tests, or other repository files.
    Effectiveness measurement: The script exits non-zero if any handoff artifact remains.
    Slice rationale: Required terminal scrub for every implementation workflow.
    Architectural effect: None; hygiene only.
    Goal: Leave the branch free of handoff artifacts.
    Motivation: Handoff files must not reach the PR.
    Alternative considerations: Manual cleanup was ruled out as non-deterministic.
    Implementation details: Run scripts/scrub-handoff-artifacts.sh.
    Non-goals: No product edits.
    Layer: docs
    Feature state: active
    Acceptance criteria:

  • bash scripts/scrub-handoff-artifacts.sh exits 0.

Revert Plan

Revert Plan
  • Safe to revert? Yes
  • Revert command: git revert <sha>
  • Post-revert steps: None
  • Data migration? No

Note

Low Risk
Documentation-only edits; no changes to install behavior, hooks, tests, or skills.

Overview
Documents the existing ./install.sh --engine-only flag so install docs and the ecosystem overview explain bare-engine installs without reading install.sh.

README adds an Engine-only mode subsection under Install: the flag links core engine skills, hooks, always-on rules, and the four cited product gates; prunes other corpus/product symlinks; points Claude at engine/CLAUDE.core.md instead of mined corpus/CLAUDE.learned.md; and notes that a full ./install.sh restores the stack while corpus remains in git and keeps growing from reflect / automate-me.

docs/ecosystem.md adds a matching Engine-only mode section, including ENGINE_CORE_PRODUCT_SKILLS, symlink pruning, the learned-rules / session-hygiene mirror behavior, and that corpus still accumulates mined output in git until a normal install relinks everything.

Reviewed by Cursor Bugbot for commit 014127f. Bugbot is set up for automated code reviews on this repo. Configure here.

…im: README.md and docs/ecosystem.md gain an "Engine-only mode" section that matches the shipped flag behavior.

Review lane: docs
Safety invariant: Text only; install.sh, tests, hooks, and skills are untouched. Confirmed by the user on 2026-09-01.
Effectiveness measurement: `grep -c -- '--engine-only' README.md` and `grep -c -- '--engine-only' docs/ecosystem.md` each print at least 1.
Slice rationale: Text is its own diff so reviewers judge wording, not code.
Architectural effect: None; prose only.
Goal: Add the two sections.
Motivation: The flag exists but nothing in the repo text mentions it.
Alternative considerations: Folding the text into the installer slice was ruled out because that mixes prose review with shell review.
Implementation details: In README.md, under the existing install section (near the `./install.sh --force` mention), add an "Engine-only mode" subsection: one short paragraph saying `./install.sh --engine-only` links only the engine (reflect, automate-me, create-skill, draft-pr, make-pr, thrash-reflect-automate, all engine hooks, always-on rules) plus the four gates the engine cites (the ENGINE_CORE_PRODUCT_SKILLS array in install.sh names them), prunes every corpus and other product symlink from the three harness skill folders, and links ~/.claude/CLAUDE.md to engine/CLAUDE.core.md; a plain `./install.sh` restores everything. Add one sentence that corpus stays in git and refills as reflect and automate-me run, so a newer model can regenerate principles from scratch. In docs/ecosystem.md add an "Engine-only mode" paragraph after the Buckets table saying the same in two or three sentences and naming the four gates from that array.
Non-goals: No code, test, hook, or skill edits; no other README restructuring.
Layer: docs
Feature state: active
Files:
- README.md
- docs/ecosystem.md
Change types:
- README.md: modify
- docs/ecosystem.md: modify
Acceptance criteria:
- `grep -q -- '--engine-only' README.md && grep -q -- '--engine-only' docs/ecosystem.md` exits 0.
- `git diff --stat main -- . ':!README.md' ':!docs/ecosystem.md'` shows no other files changed.

Exit code: 0
…aim: Both files mention the flag and nothing else changed.

Review lane: proof
Safety invariant: Read-only check.
Effectiveness measurement: The grep counts are the measurement.
Slice rationale: One focused proof for the text slice.
Architectural effect: None; verification only.
Goal: Prove both files mention the flag.
Motivation: A text change is only reviewable when its presence is machine-checked.
Alternative considerations: Manual reading was ruled out as non-deterministic.
Implementation details: Run the two greps.
Non-goals: No mutations.
Layer: docs
Feature state: active
Acceptance criteria:
- Exits 0 only when both files mention the flag.

Exit code: 0
…No ephemeral inter-task handoff files remain in the worktree before the merge gate.

Review lane: cleanup
Safety invariant: The scrub script only removes known handoff artifact names and never touches source, tests, or other repository files.
Effectiveness measurement: The script exits non-zero if any handoff artifact remains.
Slice rationale: Required terminal scrub for every implementation workflow.
Architectural effect: None; hygiene only.
Goal: Leave the branch free of handoff artifacts.
Motivation: Handoff files must not reach the PR.
Alternative considerations: Manual cleanup was ruled out as non-deterministic.
Implementation details: Run scripts/scrub-handoff-artifacts.sh.
Non-goals: No product edits.
Layer: docs
Feature state: active
Acceptance criteria:
- `bash scripts/scrub-handoff-artifacts.sh` exits 0.

Exit code: 0
…a07eb591a-9e62cdfc — Review claim: No ephemeral inter-task handoff files remain in the worktree before the merge gate.

Review lane: cleanup
Safety invariant: The scrub script only removes known handoff artifact names and never touches source, tests, or other repository files.
Effectiveness measurement: The script exits non-zero if any handoff artifact remains.
Slice rationale: Required terminal scrub for every implementation workflow.
Architectural effect: None; hygiene only.
Goal: Leave the branch free of handoff artifacts.
Motivation: Handoff files must not reach the PR.
Alternative considerations: Manual cleanup was ruled out as non-deterministic.
Implementation details: Run scripts/scrub-handoff-artifacts.sh.
Non-goals: No product edits.
Layer: docs
Feature state: active
Acceptance criteria:
- `bash scripts/scrub-handoff-artifacts.sh` exits 0.
@cursor

cursor Bot commented Sep 12, 2026

Copy link
Copy Markdown

Bugbot couldn't run - usage limit reached

Bugbot is counted against Cursor usage for this user or team, and this run hit a usage or spend limit.

A user or team admin can review and increase usage limits in the Cursor dashboard.

(requestId: serverGenReqId_96a2d5fd-e2af-4a89-8af5-4d6517d6d95c)

@EdbertChan EdbertChan closed this Sep 12, 2026
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant