Skip to content

docs: codify SDD development process (issue #390)#392

Open
konard wants to merge 2 commits into
ProverCoderAI:mainfrom
konard:issue-390-5cc8c59392dd
Open

docs: codify SDD development process (issue #390)#392
konard wants to merge 2 commits into
ProverCoderAI:mainfrom
konard:issue-390-5cc8c59392dd

Conversation

@konard

@konard konard commented Jun 10, 2026

Copy link
Copy Markdown
Contributor

Summary

Issue #390 («Как должна проходить разработка?») describes the intended
end-to-end development paradigm for docker-git. This PR codifies that
paradigm as canonical documentation so the workflow becomes verifiable and
repeatable rather than tribal knowledge.

What changed

  • docs/process.md — new doc describing the full SDD (Spec-Driven
    Development) cycle, mapping each step to existing tooling:
    • clone --mcp-playwright to provision an isolated env;
    • /plan → clarifying questions + Deep Research → Story ending with a
      Prove block of verifiable invariants;
    • approved plan persisted to the PR via
      plan-to-git (already wired
      in generated envs by feat(shell): sync agent plans to pull requests #371);
    • development via ToDos/subagents, tests and Playwright MCP;
    • CI + verification of each Story invariant.
  • AGENTS.md / CLAUDE.md — added the issue's follow-up requirements to
    the existing PROOF-obligations section:
    • mandatory proof-of-fix (why the problem existed + how it was solved +
      the test that failed before and passes after);
    • explicit ban on code/meaning duplication (Deep Research before new
      code; one unit of meaning → one definition site).
  • README.md — links the new process doc.
  • Removed the placeholder .gitkeep.

Why documentation

The issue and its comments ask "how should development proceed" and "how the
paradigm should be structured" — i.e. they request a defined, provable process.
Most of the runtime pieces already exist (--mcp-playwright, --auto,
plan-to-git, the formal-proof philosophy in AGENTS.md/CLAUDE.md); the gap
was a single canonical description tying them into one verifiable loop and the
two newly-stated requirements (proof-of-fix, anti-duplication).

Verification

Docs-only change. docs/process.md links resolve to existing files
(AGENTS.md, CLAUDE.md, README.md) and the README link points to the new
doc. No package code touched, so no changeset/version bump.

Closes #390

Adding .gitkeep for PR creation (default mode).
This file will be removed when the task is complete.

Issue: ProverCoderAI#390
@coderabbitai

coderabbitai Bot commented Jun 10, 2026

Copy link
Copy Markdown

Warning

Review limit reached

@konard, we couldn't start this review because you've reached your PR review rate limit.

More reviews will be available in 53 minutes and 58 seconds. Learn how PR review limits work.

Your organization has run out of usage credits. Purchase more in the billing tab.

⌛ How to resolve this issue?

After more reviews become available, a review can be triggered using the @coderabbitai review command as a PR comment. Alternatively, push new commits to this PR.

We recommend that you space out your commits to avoid hitting the rate limit.

🚦 How do rate limits work?

CodeRabbit enforces hourly rate limits for each developer per organization.

Our paid plans include higher PR review limits than trial, open-source, and free plans. In all cases, reviews become available again over time. During sustained high-volume PR review activity, CodeRabbit may temporarily slow when the next review becomes available.

Please see our Fair Usage Limits Policy for further information.

ℹ️ Review info
⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: ASSERTIVE

Plan: Pro Plus

Run ID: 5d4447f1-dd32-4ee5-8bae-203416c9b21a

📥 Commits

Reviewing files that changed from the base of the PR and between 8889c5c and bc011b1.

📒 Files selected for processing (4)
  • AGENTS.md
  • CLAUDE.md
  • README.md
  • docs/process.md
✨ Finishing Touches
🧪 Generate unit tests (beta)
  • Create PR with unit tests

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

Comment @coderabbitai help to get the list of available commands and usage tips.

Add docs/process.md describing the end-to-end Spec-Driven Development
cycle from issue ProverCoderAI#390: clone --mcp-playwright, /plan with clarifying
questions + Deep Research, Story with Prove (verifiable invariants),
plan-to-git PR persistence, development via ToDos/subagents with
Playwright MCP, CI and invariant verification.

Incorporate the issue's follow-up requirements into AGENTS.md/CLAUDE.md:
mandatory proof-of-fix on bug fixes and a ban on code/meaning
duplication. Link the process doc from README.

Closes ProverCoderAI#390
@konard konard changed the title [WIP] Как должна проходить разработка? docs: codify SDD development process (issue #390) Jun 10, 2026
@konard konard marked this pull request as ready for review June 10, 2026 10:24
@konard

konard commented Jun 10, 2026

Copy link
Copy Markdown
Contributor Author

Working session summary

Done. PR #392 is ready for review with green CI.

Summary

Issue #390 ("Как должна проходить разработка?") asks how development should proceed and, in its comments, requests that the paradigm be structured so work is provable. It's a process/vision issue — most runtime pieces already exist (--mcp-playwright, --auto, plan-to-git from #371, the formal-proof philosophy in AGENTS.md/CLAUDE.md); the gap was a single canonical, verifiable description plus two newly-stated rules.

Changes (docs-only):

  • docs/process.md — codifies the full SDD (Spec-Driven Development) loop, mapping each step to existing tooling: clone --mcp-playwright/plan (clarifying questions + Deep Research) → Story ending in a Prove block of verifiable invariants → approved plan persisted to the PR via plan-to-git → development with ToDos/subagents + Playwright MCP → CI + per-invariant verification.
  • AGENTS.md / CLAUDE.md — added the comment requirements to the existing PROOF-obligations section: mandatory proof-of-fix (cause + solution + the test that failed-before/passes-after) and a ban on code/meaning duplication.
  • README.md — links the new doc; removed the placeholder .gitkeep.

Verification: All 18 real CI checks pass (Build, Test, Types, Lint, Lint Effect-TS, full E2E matrix, cross-OS Final build, Snapshot). The single non-passing check is CodeRabbit reporting "Insufficient review credits" — an external billing condition, not a code failure. PR is mergeable.

PR: #392

One note for you: this issue was open-ended ("how should development proceed"), so I interpreted the deliverable as canonical documentation of the workflow rather than new runtime tooling. If you'd instead like a concrete feature (e.g. a /plan skill file, or automated invariant-checking in CI), let me know and I'll extend the PR.


This summary was automatically extracted from the AI working session output.

@konard

konard commented Jun 10, 2026

Copy link
Copy Markdown
Contributor Author

🤖 Solution Draft Log

This log file contains the complete execution trace of the AI solution draft process.

💰 Cost: $1.688603

📊 Context and tokens usage:

  • 66.1K / 1M (7%) input tokens, 16.8K / 128K (13%) output tokens

Total: (3.0K new + 87.8K cache writes + 1.4M cache reads) input tokens, 16.8K output tokens, $1.688603 cost

🤖 Models used:

  • Tool: Anthropic Claude Code
  • Requested: opus
  • Model: Claude Opus 4.8 (claude-opus-4-8)

📎 Log file uploaded as Gist (1035KB)


Now working session is ended, feel free to review and add any feedback on the solution draft.

@konard

konard commented Jun 10, 2026

Copy link
Copy Markdown
Contributor Author

🟡 Ready for review

Hive Mind stopped automatic restart because the remaining failed check is an external review quota/credit limit, not a code failure it can fix.

Checks not executed:

  • CodeRabbit — Insufficient review credits

Checks completed successfully:

Action required:

  • Restore the external review credits/rate limit and rerun the review, or decide manually whether this PR can proceed.
  • No new AI session was started for this blocker.

Monitored by hive-mind with --auto-restart-until-mergeable flag.

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.

Как должна проходить разработка?

1 participant