v2.9.0 — Batch Discipline

Scaling up SmartWorker batch and unattended work safely: a tier must earn
each work type before it runs at scale, parallel dispatch gets hard rules,
and no subtask ever runs twice.

Added

• SmartWorker calibration gate — tier / capability labels are now
  framed as execution slots, not permanent truths. Before a tier is
  trusted with a new class of work (especially batch/unattended runs), a
  small calibration pilot must pass the owner's quality threshold, recorded
  in the spec's new calibration field and re-run when the underlying model
  changes. The capability twin of the mechanical-gate rule: a tier is not
  suitable until a pilot shows it passing.
• SmartWorker parallel fan-out — the Main Agent may dispatch multiple
  workers concurrently for many same-shaped independent subtasks (new
  optional concurrency spec field) — star topology under the single Main,
  never a peer mesh. Five rules: calibrate first; owner-set concurrency cap
  • provider rate limits; disjoint write targets per worker; a per-batch
    (not per-worker) budget ceiling; and single assignment + idempotent
    outputs (dispatch manifest, retry only confirmed-failed items, reconcile
    against real artifacts on resume, prefer overwrite-by-key over appends —
    no subtask runs twice). Applies equally when a Loop Contract's run fans
    out internally.
• Loop Contract duplicate-work guard — the cross-run state anchors are
  now explicitly the duplicate-work guard: a worker checks actual artifact
  state before doing an item and never re-does what the anchors show as
  done; the definition-of-done check must be idempotent so a retried or
  re-entered run converges instead of repeating work.

Changed

• Loop Contract template: Contract Restate moved into the run report —
  the kickoff prompt's "restate before doing anything" instruction was
  skipped in live trial runs (chat-level asks are decorative); the restate
  is now a required opening section of the run report itself, making it
  part of the machine-checkable done condition.

Notes

Every feature in this release was motivated and validated by live
downstream usage before shipping — real batch knowledge-work runs and the
governed autonomous-run trials behind v2.8.0.

Upgrading

Use the AI-driven upgrade prompt in the README ("Upgrading an existing
workspace") — preserve-first: your filled context files and records are
never overwritten.

Full changelog: CHANGELOG.md · v2.8.0...v2.9.0

v2.8.0 — Governed Autonomous Runs

AgentJoJoy can now govern unattended agent runs: a SmartWorker running on a
schedule (or kicked off and left alone) operates under an owner-approved Loop
Contract — with every rule in this release validated live on a real project
before shipping.

Added

• Loop Contracts (AgentJoJoy/agent-smartworkers/loop-contract.template.md) —
  run a SmartWorker unattended under a contract: machine-checkable definition of
  done, work tiers (Run-to-Done / Deliver-to-Gate / never-autonomous), strict
  write scope, budget & pacing caps, cross-run no-progress rule, and escalation.
  The owner approves the envelope once; no human sits in the loop. Activation
  requires mechanically enforced gates proven by a live blocking pre-flight.
  AgentJoJoy ships no scheduler — your runtime's native cron executes; the
  contract governs.
• Hermes Agent runtime binding (AgentJoJoy/agent-hooks/runtime-bindings/hermes.md) —
  how to enforce gates on Hermes: plugin vs shell-hook vs gateway-hook surfaces,
  the activation chains that fail open and silently, decision mapping, and
  live-observed caveats.

Changed

• Sharper positioning — AgentJoJoy is the context and governance layer
  between the language model and the execution runtime; it does not reason and
  does not execute.
• "A task instruction is not approval" — closed a live-observed gap where an
  agent treated task wording ("fix it, then commit") as the approval itself; the
  show-the-exact-command-and-wait step always applies.
• Root namespace handoff now covers CHANGELOG.md/LICENSE — after
  onboarding they move under AgentJoJoy/, leaving the workspace root to your
  project.

Fixed

• Packaged files no longer carry a UTF-8 BOM — some runtimes block context
  files containing invisible unicode (observed live: AGENTS.md silently not
  loading on Hermes). A new packaging check fails any release that ships a BOM.
• Rule-consistency fixes: CLAUDE.md Critical Rules reference, SPEC-4.5 vs
  milestone-commit carve-out, engagement-mode checkbox citation.

Validated before adoption

Every behavior in this release was proven on a live downstream workspace first:
two full autonomous runs (a research worker and a test worker driving a real
headless backtest), an owner curation gate between them, and a mechanical git
gate verified by a live blocking pre-flight — including an honestly-reported
losing run ("Rejected for promotion").

Upgrading

Use the AI-driven upgrade flow in the README (point your agent at the canonical
upgrade prompt); your trackers, decisions, and custom skills are preserved.

Full changelog: CHANGELOG.md ·
Compare: v2.7.0...v2.8.0

v2.7.0 — Surface Lifecycle & Wrapper Handoff

This release clarifies AgentJoJoy's lifecycle model so runtime context stays lean, completed setup state can archive cleanly, and the real project folder remains the center of project git/documentation.

Changed

• Clarified setup-tracker.md as temporary setup/onboarding state with setup-history/ for completed setup history.
• Classified SmartWorkers, Work Item Envelopes, communication behavior surfaces, optional project context, hooks, templates, and incubator surfaces by lifecycle role.
• Clarified root README.md as bootstrap scaffolding only; after onboarding, project documentation belongs in <project-folder>/README.md.

Upgrading

Use the canonical AI-driven upgrade prompt in the public README.

## v2.6.1 — Preserve-First Unwrap

What's Changed

• Replaced the old clean-ejection flow with preserve-first unwrapping.
  Instead of running a bundled delete script, users now ask the current Main Agent to inventory AgentJoJoy files, identify what should be preserved, show a dry-run cleanup plan, and wait for explicit approval before deleting anything.

• Removed the bundled ejection helper and the empty AgentJoJoy/agent-tools/ public surface.
  AgentJoJoy stays a governance/document layer; one-off cleanup scripts can still be created by the Main Agent when they fit the workspace.

• Fixed the v2.6.0 changelog release date so it matches the actual shipped date.

Compatibility Note

The old ejection command is no longer shipped. To unwrap AgentJoJoy, use the Main Agent-guided flow documented in the README.

Verification

• Version Alignment: PASS
• Safety Shield: PASS
• Broken-link scan: 0 / 131
• Active public docs leak scan: clean

PR: #5
Compare: v2.6.0...v2.6.1

v2.6.0 — Agent Records & Execution-Mode Governance

AgentJoJoy v2.6.0 consolidates workspace records under AgentJoJoy and tightens execution-mode governance while keeping the framework docs-only and Main-Agent-first.

Added

• Public README visual hero.
• Optional work records archive under AgentJoJoy/agent-records/work/.
• Work Item Envelope template for selected completed or paused work.

Changed

• Execution-mode choice discipline now defaults to the single Main Agent and treats over-surfacing as an anti-pattern.
• Progress tracker, setup tracker, decision guide, and work records now live under AgentJoJoy/agent-records/.

Fixed

• Progress tracker relative links after the records move.
• Removed the empty legacy AgentJoJoy/agent-runtime/ package surface and active public references.

Upgrading

Use the upgrade prompt in the public README and preserve existing agent-context/ and agent-records/ content.

v2.5.0 — Governance-Layer Positioning & Local Provider SmartWorkers

This release clarifies AgentJoJoy’s product boundary and adds optional SmartWorker fields for local or alternative-provider workers without turning AgentJoJoy into an agent runtime.

Added

• SmartWorker local / alternative-provider extension: smartworker-spec.template.md now includes optional owner-controlled fields for provider / runtime, model-capability assumption, privacy, and cost policy.
• Local-provider SmartWorkers stay opt-in and docs-only: no provider config, installer, model recommendation, daemon, or active runtime ships by default.
• Capability labels such as local-light, local-code, local-reasoning, and local-embed are used instead of model catalogs.

Changed

• Public README positioning now states the core boundary directly: AgentJoJoy is a governance layer, not an agent runtime.
• Frontier Main remains the controller and reviewer. Local-worker output is treated as untrusted draft, not a safety boundary.

Upgrading

Use the canonical AI-driven upgrade prompt in the public README.

Full changelog: https://github.com/Joyperm/AgentJoJoy/blob/v2.5.0/CHANGELOG.md
Compare: v2.4.1...v2.5.0

v2.4.1 — Rule Integrity & Intake-Driven Attribution

This patch release tightens the public rule surface after the Hook Enforcement Contracts release and makes commit attribution a repo/intake policy instead of a universal AI trailer rule.

Changed

• Commit attribution is now intake-driven: AgentJoJoy records the repo's commit attribution policy in standards.md, scans Path 2 team rules/recent commits before asking, and defers multi-agent snippets to the target repo convention unless AI co-author trailers are explicitly enabled.

Fixed

• Public package rule-surface cleanup: removed internal template-development residue from public-facing docs so copied workspaces only see the normal public workflow.

Validation

• Markdown relative links checked: 0 broken out of 130
• Public package checked for dev-only markers, local paths, and source-repo residue

Upgrading

Use the AI-driven upgrade prompt in the public README to compare and merge the new template guidance into an existing workspace.

v2.4.0 — Hook Enforcement Contracts

This release adds optional Hook Enforcement Contracts for owners who want to mechanize selected AgentJoJoy or project-specific gates without enabling any active hooks by default.

Added

• Added AgentJoJoy/agent-hooks/ as a docs/template-only layer.
• Added a blank neutral Hook Enforcement Contract template.
• Added runtime binding notes for Claude Code, Codex, and Antigravity.
• Documented opt-in install scope, owner-defined hooks, runtime enforcement posture, and hook safety guardrails.

Notes

• No hook scripts, hook configs, installers, CI checks, git hooks, active enforcement, or default enablement are included.
• Existing AgentJoJoy gates remain sourced in agent-rules/ / workflow-spec.md.
• agent-hooks/ only maps optional enforcement contracts back to canonical rules.

Validation

• release.ps1 Version Alignment PASS.
• release.ps1 Safety Shield PASS.
• Broken-link scan: 0 / 130 local links.
• Public diff checked: expected files only; no dev-only directories or new leak patterns.

Upgrading

Use the AI-driven upgrade flow in the public README so your local project notes, decisions, and custom skills are preserved.

Full changelog: https://github.com/Joyperm/AgentJoJoy/blob/v2.4.0/CHANGELOG.md
Compare: v2.3.0...v2.4.0

v2.3.0 — SmartWorker Framework

This release adds AgentJoJoy’s SmartWorker framework: a single-Main-Agent pattern for dispatching knowledge-heavy recurring work into a separate context without turning AgentJoJoy into multi-agent orchestration.

Added

• SmartWorker framework under AgentJoJoy/agent-smartworkers/.
• Runtime-neutral SmartWorker spec template.
• Public packaging for agent-smartworkers/.

Changed

• pattern-detection now routes repeated work to script / SmartWorker / Skill.
• Worker escalation taxonomy is consolidated in ai-workflow-rules.md.
• Entry points, workflow guides, and skills docs now explain SmartWorker dispatch and clarify that SmartWorkers are not Skills.
• Release-facing docs now use the current read-only Resume Check wording instead of stale Worktree Auto-Sync language.

Validation

• Version Alignment PASS.
• Safety Shield PASS.
• Broken-link scan: 0 / 117.
• No dev-only dirs, markers, or path leaks.

Upgrading

Use the canonical AI-driven upgrade prompt in the public README.

Full changelog: https://github.com/Joyperm/AgentJoJoy/blob/v2.3.0/CHANGELOG.md
Compare: v2.2.0...v2.3.0

v2.2.0 — Configurable Engagement & Lean Output

This release makes AgentJoJoy’s engagement behavior more explicitly configurable while adding a concise-output mode that preserves reasoning and safety.

Added

• Lean Output mode + lean-output skill: an opt-in brevity style for shorter delivery without reducing reasoning, code correctness, teaching substance, or approval/safety text.

Changed

• engagement-mode.md now has an explicit Autonomy Configuration surface with owner-controlled Teaching box and Lean Output toggles.
• Pillar II teaching-box wording is reconciled so teaching is toggle-controlled rather than always-on.
• Safety gates and efficiency guards remain always-on and are not toggleable.

Validation

• release.ps1: Version Alignment PASS, Safety Shield PASS.
• Broken-link scan: 0 broken links across 108 local relative links.
• Public packaging confirmed no dev-only dirs/markers/path hits.

Upgrading

Use the canonical AI-driven upgrade prompt in the public README.

Full changelog: https://github.com/Joyperm/AgentJoJoy/blob/v2.2.0/CHANGELOG.md
Compare: v2.1.1...v2.2.0