Skip to content

docs(release): prepare v0.5.0 notes - #296

Merged
egdev6 merged 63 commits into
mainfrom
release/v0.5.0
Jul 24, 2026
Merged

egdev6 merged 63 commits into
mainfrom
release/v0.5.0

Conversation

@egdev6

@egdev6 egdev6 commented Jul 23, 2026

Copy link
Copy Markdown
Owner

Closes #210

PR Type

  • Bug fix
  • New feature
  • Documentation only
  • Code refactoring
  • Maintenance/tooling
  • Breaking change

Summary

  • Prepares sanitized v0.5.0 release notes for the develop-to-main promotion.
  • Fixes release changeset generation so the final unterminated git log --pretty=format row is preserved.
  • Adds a safe validator for final-row handling and token/secret redaction in generated release notes.

Changes

File Change
docs/releases/v0.5.0.md Adds v0.5.0 release notes and validation checklist.
scripts/generate-release-changeset.sh Preserves final git-log rows and broadens named release-note redaction patterns.
scripts/validate-release-changeset-final-row.sh Adds deterministic fake-repo validation for final-row and redaction behavior.

Test Plan

  • bash scripts/validate-release-promotion.sh --base origin/main --head HEAD --force
  • bash scripts/validate-release-changeset-final-row.sh
  • bash scripts/validate-repo-safe-evidence.sh
  • bash -n scripts/generate-release-changeset.sh scripts/validate-release-changeset-final-row.sh
  • git diff --check

Contributor Checklist

  • Linked an approved issue
  • Added exactly one type:* label
  • Ran syntax checks on modified scripts
  • Docs updated for release behavior
  • Conventional commit format
  • No Co-Authored-By trailers

egdev6 added 30 commits June 9, 2026 17:32
* ci: harden stable merge checks

* ci: format evidence guard script
* feat(runtime): package Gentle-AI for OpenClaw

* ci: satisfy Dockerfile lint for Gentle-AI install
* refactor(runtime): curate OpenClaw skill surface

* fix(runtime): satisfy shellcheck for skill sync pruning
* ci(release): automate changesets and tagging

* style(release): format release automation scripts
* docs(runtime): define artifact classification and private data policy

* docs(runtime): clarify artifact classification validation

* test(runtime): cover workflow skill classification
* test(runtime): record private Discord Engram live preflight

* test(runtime): clarify live preflight approval blocker

* test(runtime): fix live preflight shell lint
* test(runtime): prove synthetic Discord no-op observation

* test(runtime): reject token-like no-op inputs

* test(runtime): fix noop shell lint

* test(runtime): format noop validator
* test(runtime): add Discord runtime boundary harness

* test(runtime): harden boundary harness checks

* test(runtime): cover noop memory write boundary

* test(runtime): fix boundary harness shell lint
* test(runtime): prepare private no-op rehearsal evidence

* test(runtime): enforce rehearsal safety gates

* test(runtime): fix rehearsal shell lint
* docs(release): record private noop verification

* docs(runtime): clarify noop rehearsal result

* docs(runtime): align noop rehearsal guide

* docs(release): fix issue reference lint

* chore(ci): retrigger release note checks
egdev6 added 17 commits June 20, 2026 18:16
* test(runtime): prepare private noop result packet

* fix(ci): format noop result validator
* chore(release): promote v0.2.0 runtime governance baseline (#241)

* chore(release): promote v0.3.0 Discord admin baseline (#251)

* docs(release): prepare v0.4.0 notes (#285)

* ci: harden stable merge checks (#165)

* ci: harden stable merge checks

* ci: format evidence guard script

* docs(repo): align templates for release readiness (#164)

* feat(runtime): package Gentle-AI for OpenClaw (#167)

* feat(runtime): package Gentle-AI for OpenClaw

* ci: satisfy Dockerfile lint for Gentle-AI install

* fix(runtime): package Engram CLI for OpenClaw MCP (#169)

* docs(runtime): document OpenAI OAuth setup (#171)

* docs(discord): add OpenClaw Global channel guides (#172)

* feat(discord): add safe project delete contract (#173)

* docs(ci): align workflow trigger docs (#178)

* test(runtime): document OpenClaw Gentle orchestration validation (#179)

* refactor(runtime): curate OpenClaw skill surface (#180)

* refactor(runtime): curate OpenClaw skill surface

* fix(runtime): satisfy shellcheck for skill sync pruning

* ci(release): automate changesets and tagging (#184)

* ci(release): automate changesets and tagging

* style(release): format release automation scripts

* docs(runtime): define governance foundation (#215)

* docs(runtime): define artifact classification and private data policy

* docs(runtime): clarify artifact classification validation

* test(runtime): cover workflow skill classification

* feat(runtime): classify Discord request artifacts (#216)

* feat(runtime): resolve context and audit durable changes (#217)

* feat(runtime): define private runtime backup restore (#218)

* chore(runtime): pin validated runtime versions (#219)

* test(runtime): add private Discord Engram evidence contract (#220)

* test(runtime): cover approval gate lifecycle (#221)

* test(runtime): add private Discord Engram readiness gate (#222)

* test(runtime): design private Discord Engram no-op observation (#223)

* test(runtime): define approval enforcement repair contract (#224)

* test(runtime): sync private Discord Engram readiness gate (#225)

* test(runtime): record private Discord Engram live preflight (#226)

* test(runtime): record private Discord Engram live preflight

* test(runtime): clarify live preflight approval blocker

* test(runtime): fix live preflight shell lint

* feat(runtime): add Discord approval guard CLI (#229)

* test(runtime): prove synthetic Discord no-op observation (#230)

* test(runtime): prove synthetic Discord no-op observation

* test(runtime): reject token-like no-op inputs

* test(runtime): fix noop shell lint

* test(runtime): format noop validator

* test(runtime): add Discord runtime boundary harness (#232)

* test(runtime): add Discord runtime boundary harness

* test(runtime): harden boundary harness checks

* test(runtime): cover noop memory write boundary

* test(runtime): fix boundary harness shell lint

* test(runtime): prepare private no-op rehearsal evidence (#234)

* test(runtime): prepare private no-op rehearsal evidence

* test(runtime): enforce rehearsal safety gates

* test(runtime): fix rehearsal shell lint

* docs(release): record private noop verification (#236)

* docs(release): record private noop verification

* docs(runtime): clarify noop rehearsal result

* docs(runtime): align noop rehearsal guide

* docs(release): fix issue reference lint

* chore(ci): retrigger release note checks

* docs(release): clarify QA-07 gate status (#237)

* docs(release): finalize v0.2.0 notes (#238)

* docs(skills): add release roadmap hygiene workflow (#243)

* feat(skills): define OpenClaw skill development handoff (#244)

* feat(discord): add general advisor contract (#245)

* feat(discord): define admin UX contract (#247)

* test(discord): add admin UX preview harness (#249)

* test(discord): add admin UX preview harness

* fix(docs): satisfy admin harness markdown lint

* test(runtime): add private write readback preflight gate (#253)

* test(runtime): prove approval enforcement boundary (#255)

* test(runtime): prove noop observation boundary (#257)

* test(runtime): sync write readback readiness proofs (#259)

* test(runtime): prove proposal binding boundary (#261)

* test(runtime): prepare private topology readiness packet (#263)

* test(runtime): prepare private topology readiness packet

* style(scripts): format topology readiness validator

* chore(ci): trigger topology readiness checks

* test(runtime): prepare redacted noop ingestion packet (#265)

* test(runtime): prepare private execution approval packet (#267)

* test(runtime): prepare noop ingestion review packet (#269)

* test(runtime): add noop ingestion review harness (#271)

* test(runtime): sync noop ingestion harness readiness (#273)

* test(runtime): prepare private noop runbook gate (#275)

* test(runtime): sync private noop runbook readiness (#277)

* test(runtime): prepare private no-op execution result packet (#279)

* test(runtime): prepare private noop result packet

* fix(ci): format noop result validator

* chore(runtime): update Gentle-AI baseline to 2.1.11 (#282)

* chore(runtime): update Engram baseline to v1.20.0 (#284)

* docs(release): prepare v0.4.0 notes

* fix(release): avoid shellcheck warning in changeset template

* ci: ignore merge commits in commitlint ranges
* feat(content): add source ingestion contract

* style(scripts): format content ingestion validator

* style(fixtures): format content ingestion fixture
egdev6 added 2 commits July 23, 2026 22:57
# Conflicts:
#	docs/architecture/discord-general-advisor.md
#	docs/architecture/discord-runtime-orchestrator.md
#	docs/releases/v0.4.0.md
#	examples/discord-approval-gate.fake.yaml
#	examples/discord-general-advisor.fake.yaml
#	examples/discord-runtime-orchestrator.fake.yaml
#	scripts/generate-release-changeset.sh
#	scripts/validate-discord-approval-gate.sh
#	scripts/validate-discord-general-advisor.sh
#	scripts/validate-discord-runtime-orchestrator.sh
#	skills/discord-approval-gate/SKILL.md
#	skills/discord-general-advisor/SKILL.md
#	skills/openclaw-runtime-orchestrator/SKILL.md
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Projects

None yet

Development

Successfully merging this pull request may close these issues.

release: promote v0.5.0 from develop to main

1 participant