Skip to content

gstack2: inferred execution profiles + capability readiness + CI hardening#12

Closed
time-attack wants to merge 8 commits into
codex/gstack-2from
time-attack/gstack2-readiness-and-ci-hardening
Closed

gstack2: inferred execution profiles + capability readiness + CI hardening#12
time-attack wants to merge 8 commits into
codex/gstack-2from
time-attack/gstack2-readiness-and-ci-hardening

Conversation

@time-attack

@time-attack time-attack commented Jul 21, 2026

Copy link
Copy Markdown
Owner

Summary

Combined GStack 2 branch bundling three independent pieces of work onto codex/gstack-2. No version bump or CHANGELOG entry — this is intended to fold into a larger change later.

Inferred execution profiles (feat: infer execution profiles)

  • Shared readiness / standard / deep depth inference across all six skills, keyed on structured operating conditions (risk, blast radius, irreversibility, mutation authority, deployment/release state, evidence need, scope) — never prompt keywords.
  • Generated EXECUTION-PROFILES.md contract per skill defining mandatory modules, legal skips, artifacts, and allowed claims. Readiness must state "Readiness profile — not a complete review." and cannot claim comprehensive/verified/production-safe results, so smoke can never masquerade as a complete review.
  • Wired through the dispatcher preamble, routing types, host schema, and parity fixtures.

Capability readiness process (feat: add capability readiness process, fix: treat hard runtime failure as not-ready)

  • gstack doctor capability readiness axes plus runtime/doctor.js, RUNTIME.md guidance across the six skills, and docs/gstack-2/CAPABILITY-READINESS.md.
  • Hard runtime failure now resolves to not-ready rather than a false green.

GitHub Actions hardening (chore(security): harden GitHub workflows)

  • Pinned/permissioned workflows plus OpenSSF Scorecard, OSV-Scanner, and dependency-review jobs. Independent of the two GStack 2 changes (only .github/workflows/ + one test).

Test Coverage

  • bun run test:gstack2: 226 pass, 0 fail (includes routing, UX, parity, and capability-readiness suites).
  • bun test test/release-hardening.test.ts: 11 pass, 0 fail.
  • Free bun test suite: exit 0.
  • Generation is deterministic: bun run gen:gstack2 produces no diff on the combined tree.

Pre-Landing Review

No issues found. Changes are additive: pure inference functions + deterministic generated contracts + runtime readiness axes + CI workflow files. No SQL/data-safety or LLM-trust-boundary surface; no frontend. The two GStack 2 features share one generator file (scripts/gstack2/generate-skill-tree.ts) and auto-merged on separate lines.

Notes

  • Six public skills preserved; no seventh skill.
  • Base: codex/gstack-2. No version bump / CHANGELOG by request.

Summary by cubic

Adds inferred execution profiles and a capability readiness flow to GStack 2, and hardens CI/security to prevent false greens and improve safety. This standardizes depth across all skills, upgrades gstack doctor, and pins security workflows.

  • New Features

    • Inferred execution profiles: shared readiness/standard/deep depth across all six skills; generated per-skill EXECUTION-PROFILES.md; SKILL headers now say “Depth: <readiness, standard, or deep>” with binding claims/skip rules.
    • Capability readiness: gstack doctor --capability browser|design|diagram|pdf|ios returns non‑mutating readiness with separate axes (judgment, platform, preview consent, install consent, readiness); adds human and JSON output; iOS marked unsupported off macOS; docs at docs/gstack-2/CAPABILITY-READINESS.md.
    • CI/security: adds pinned actions/dependency-review-action, weekly google/osv-scanner-action, and ossf/scorecard-action with github/codeql-action/upload-sarif; enforces least-privilege and workflow concurrency.
  • Bug Fixes

    • Hard runtime failure now reports not-ready (ok:false/exit 1) in capability readiness, matching plain gstack doctor.

Written for commit 1e5071a. Summary will update on new commits.

Review in cubic

Sinabina and others added 8 commits July 21, 2026 13:03
Add public-infrastructure security baseline for GStack 2 CI/release:
- Dependency Review on PRs to main (fail-on-severity: high)
- Weekly OSV vulnerability scan with SARIF upload
- OpenSSF Scorecard (weekly + push), publish_results disabled
- Concurrency controls on the four workflows that lacked them
Preserves immutable SHA action pins and least-privilege permissions.
Extends release-hardening.test.ts to pin these invariants.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
…/Add-capability-readiness-process-2

# Conflicts:
#	scripts/gstack2/generate-skill-tree.ts
#	skills/debug/references/RUNTIME.md
#	skills/design/references/RUNTIME.md
#	skills/plan/references/RUNTIME.md
#	skills/qa/references/RUNTIME.md
#	skills/review/references/RUNTIME.md
#	skills/ship/references/RUNTIME.md
…eadiness

A capability whose binary launches while the managed runtime hard-fails
(e.g. skill-API mismatch) was reported as `degraded`/ok:true/exit 0, diverging
from plain `gstack doctor` (ok:false/exit 1) for the identical report. Split
the branch so runtime `warn` stays `degraded` and runtime `fail` maps to
`failed`, and add a regression test for the runtime-fail + capability-pass case.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
…me-attack/gstack2-readiness-and-ci-hardening
…-attack/gstack2-readiness-and-ci-hardening
@cursor

cursor Bot commented Jul 21, 2026

Copy link
Copy Markdown

Bugbot is not enabled for your account, so this pull request was not reviewed.

Enable Bugbot in the Cursor dashboard to get automatic reviews on future PRs.

@time-attack

Copy link
Copy Markdown
Owner Author

Superseded by #10. Profiles + capability readiness + CI/security hardening are all in the consolidated PR, which also carries the execution-result contract and dependency-security upgrades this PR lacked. Closing to keep a single mergeable PR into codex/gstack-2.

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