gate-l3-lifecycle-integration
Internal quality gate that evaluates whether Hivemind harness implementations correctly participate in the runtime lifecycle — covering 9-surface mutation…
Skill by shynlee04
Install
Works with Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf. Read the repository before you give any tool access to your data.
npx agentmarket install skill/shynlee04-gate-l3-lifecycle-integration
Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.
Get the weekly diff
The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.
About
Use cases: security
Reviews
Write a reviewNo reviews yet. Installed it? Reviews come from verified installers only. How reviews work
Permissions
What this skill can do and what it asks for, from its SKILL.md. Checked Oct 25, 2026.
- Tools
- No tools found
- Secrets
- None asked for
- Install scripts
- None found
- Runs as
- Shell scripts
Network hosts it talks to 1
codeload.github.com· found in install command
Files it touches
~/.claude/skills/gate-l3-lifecycle-integration
What's inside
- Skill name
gate-l3-lifecycle-integration- Description
- Internal quality gate that evaluates whether Hivemind harness implementations correctly participate in the runtime lifecycle — covering 9-surface mutation authority, CQRS boundaries, actor hierarchy, event-driven wiring, classification fit (src/ vs .opencode/ vs .hivemind/), and OpenCode SDK surface compliance. Synthesized from .planning/codebase/ARCHITECTURE.md (9-surface authority table) and ingested @opencode-ai/plugin SDK v1.14.28 docs (tool(), hook() signatures). Use when performing a lifecycle gate check, auditing harness module integration, verifying CQRS boundary compliance, checking delegation hierarchy constraints, evaluating tool/hook registration correctness, running a harness quality gate, validating plugin composition integrity, or running phase audit on src/ modules. Activates during code review of src/ files, phase audit, milestone verification, integration check, and deployment readiness workflows.
- License
- MIT
- Metadata
- role: domain-execution · layer: 2 · pattern: P2 · version: 2.0.0 · triad-flow: lifecycle → spec compliance → evidence truth · classification: internal-quality-gate · triad-position: entry · synthesis-source: ARCHITECTURE.md + @opencode-ai/plugin SDK docs
- Instructions
- 177 lines in SKILL.md · view on GitHub
Before installing
- Ships executable scripts
Agent Skills spec issues
- allowed-tools: must be a space-separated string
Files (16)
- SKILL.md
- evals.json · evals/
- .gitkeep · metrics/
- rich-gate-scorecard.md · metrics/
- adopted-patterns.md · references/
- anti-patterns.md · references/
- cqrs-boundaries.md · references/
- evaluation-checklist.md · references/
- gap-documentation.md · references/
- nine-surface-authority.md · references/
- perspective-rubrics.md · references/
- remediation-paths.md · references/
- sdk-compliance.md · references/
- triad-flow.md · references/
- run-gate-eval.sh · scripts/
- gate-report.md · templates/
SKILL.md (first lines)
# Gate: Lifecycle Integration Internal quality gate for Hivemind harness architecture compliance. Not for end-user shipping. Activates during every gatekeeping workflow related to harness or OpenCode integration work. ## Activation Detection Load this skill when: - Code review of files in `src/` (any subdirectory) - Phase audit or milestone verification touching harness modules - Integration check after tool/hook/delegation changes - Deployment readiness for `hivemind` npm package - Any workflow referencing `gate-lifecycle-integration` ## Do NOT Load Skip when working on `.opencode/` soft meta-concept authoring, end-user feature development, non-Hivemind code reviews, documentation, or artifacts in `src/shared/`/`src/schema-kernel/` (use lighter classification check instead). ## Two-Halves Classification (Q6) Every artifact must land in exactly one of three roots: | Root | Contents | State Authority | |------|----------|-----------------| | `src/` (Hard Harness) | Tools, hooks, plugin, shared, lib | Writes to `.hivemind/` via managers only | | `.opencode/` (Soft Meta-Concepts) | Skills, agents, commands, rules | No persistent state — OpenCode primitives only | | `.hivemind/` (Deep Module State) | Journals, continuity, delegation records | Canonical state root (Q6) | Cross-contamination between roots is a **BLOCK** finding. Tools never write to `.opencode/`. Hooks never write to `.hivemind/` directly — they route through `DelegationManager` or `continuity.ts`. ## 9-Surface Mutation Authority The architecture defines 9 surfaces across write-side (4), read-side (3), and assembly (1). Each artifact must conform to its surface's authority boundaries. > **Full table with constraints**: `references/nine-surface-authority.md` > > Source: `.planning/codebase/ARCHITECTURE.md` § "9-Surface Mutation Authority" Quick summary — write-side: `continuity.ts`, `delegation-persistence.ts`, `session-journal.ts`, `DelegationManager`. Read-side: hooks, tools, sidecar. Assembly: `plugin.ts`. ## OpenCode SDK Surface Compliance Validate against the real `@opencode-ai/plugin` v1.14.28 API surface. Three areas: 1. **tool() factory**: `description`, `args` (Zod), `execute` → string. Registered in plugin.ts. 2. **Hook handlers**: Exactly 4 hooks (`tool.execute.before/after`, `experimental.session.compacting`, `shell.env`). 3. **Plugin composition**: Async function, type-only imports, no inline business logic, lazy PTY. > **Full checklists with real signatures**: `references/sdk-compliance.md` > > Additional context: `stack-opencode` skill for broader SDK reference. ## CQRS Boundary Enforcement Write-side (tools) mutate state via managers. Read-side (hooks) observe events. Events flow write→read, never reverse. 7 BLOCK-level anti-patterns detect violations (e.g., `AP-WRITE-FROM-READ`, `AP-CROSS-ROOT-WRITE`, `AP-BYPASS-MANAGER`). > **Full write/read checklists + anti-pattern table**: `references/cqrs-boundaries.md` > > Expanded anti-pattern catalog: `references/anti-patterns.md` ## Delegation Hierarchy Constraints Validate against runtime constants from `src/lib/types.ts`: | Constant | Value | Meaning | |----------|-------|---------| | `MAX_DELEGATION_DEPTH` | 3 | Max nested delegation depth | | `MAX_DESCENDANTS_PER_ROOT` | 10 | Max child delegations per root session | | `STABLE_POLLS_REQUIRED` | 3 | Consecutive unchanged polls for completion | | `TASK_CLEANUP_DELAY_MS` | 600000 | Grace period before cleanup (10 min) | Check: uses `DelegationManager.dispatch()`, valid category, depth ≤ 3, WaiterModel dispatch, dual-signal completion, recovery guarantee via `recoverPending()`. ## Decision Tree ``` START → Classify the artifact by file location: ├─ src/tools/*.ts → TOOL: tool() registration, Zod, response envelope, │ SDK mutation via session-api.ts, state via continuity.ts, LOC < 200 ├─ src/hooks/*.ts → HOOK: factory pattern, real SDK signature, CQRS readonly ├─ src/lib/*.ts → LIBRARY: dependency ≤ 2 levels, LOC < 500, no `any`, tests ├─ src/plugin.ts → COMPOSITION: LOC < 200, all registered, no inline logic ├─ src/shared/*.ts, src/schema-kernel/*.ts → LEAF: cross-cutting utility └─ DELEGATION participant? → DelegationManager, category, depth, dual-signal ``` For each branch, execute the detailed checklist in `references/evaluation-checklist.md`. ## Self-Correction ### Mode 1: When Classification Is Ambiguous If a file straddles two classification roots (e.g., a test helper in `src/shared/` that also reads `.hivemind/` state), classify by primary purpose. Helpers that are pure utility = LEAF (src/shared/). Helpers that read persistent state = suspect — check if they route through a manager. Never classify as LEAF to bypass the lifecycle gate. ### Mode 2: When CQRS Boundary Is Fuzzy Some tools produce side effects that look like state reads (e.g., delegation-status.ts reads delegation records). This is correct CQRS: the tool reads through a query API, not by subscribing to events. Distinguish: direct event subscription in a tool = BLOCK; query API call in a tool = PASS. ### Mode 3: When Plugin.ts Exceeds LOC Limit If plugin.ts exceeds 200 LOC but the excess is purely registration boilerplate (many tools/hooks), document the finding as WARNING rather than BLOCK. If the excess includes inline business logic, BLOCK. Registration-only LOC inflation is acceptable; logic inflation is not. ### Mode 4: When Delegation Depth Is At Limit A delegation chain at exactly MAX_DELEGATION_DEPTH (3) is PASS — the limit is inclusive. However, if depth=3 and the chain also uses queue keys without buildDelegationQueueKey(), flag as WARNING. At-limit depth with correct queue key construction = PASS. At-limit depth with manual key construction = WARNING (fragile). ## Gate Orchestrator Integration
From shynlee04/hivemind-plugin, MIT.
More from shynlee04/hivemind-plugin
- 0.578score
hm-l3-synthesis
shynlee04
Compress research findings into actionable artifacts with tiered reduction.
SkillCFreshness: C★ 51Claude Code +9 - 0.578score
hm-l2-coordinating-loop
shynlee04
Coordinate multi-agent dispatch with validation gates and handoff protocols.
SkillCFreshness: C★ 51Claude Code +9 - 0.578score
hm-l3-integration-contracts
shynlee04
Formal bidirectional contract authority between skills and agents.
SkillCFreshness: C★ 51Claude Code +9 - 0.578score
hf-l2-skill-synthesis
shynlee04
Use when synthesizing skills from GitHub repos, classifying skill patterns, building template libraries, generating eval frameworks, or creating eval-driven…
SkillCFreshness: C★ 51Claude Code +9 - 0.578score
hf-l2-agent-composition
shynlee04
Teaches agents how to compose specialist agents from user intent using XML markup grammar, step protocols, and structured return formats.
SkillCFreshness: C★ 51Claude Code +9 - 0.578score
hm-l3-detective
shynlee04
Investigate codebases with SCAN, READ, and DEEP reading modes.
SkillCFreshness: C★ 51Claude Code +9 - 0.574score
hf-l2-context-absorb
shynlee04
Multi-wave swarm protocol for absorbing dense context (links, text, files, stories, events, actors) into persistent session-context-prompt.md.
SkillCFreshness: C★ 51Claude Code +9 - 0.574score
hf-l2-command-parser
shynlee04
This skill should be used only when parsing $ARGUMENT propositional commands from OpenCode command strings.
SkillCFreshness: C★ 51Claude Code +9 - 0.574score
hm-l2-test-driven-execution
shynlee04
Use when executing test-first implementation, RED/GREEN/REFACTOR cycles, failing-test validation, coverage claim verification, or test-suite quality work with…
SkillCFreshness: C★ 51Claude Code +9 - 0.574score
hf-l2-skill-router
shynlee04
Maps meta-builder task domains to hf-* skill loading bundles with priority ordering.
SkillCFreshness: C★ 51Claude Code +9 - 0.574score
hm-l2-completion-looping
shynlee04
Guardrail workflows against regression with non-completion detection and automatic loop-back.
SkillCFreshness: C★ 51Claude Code +9 - 0.574score
hf-l2-delegation-gates
shynlee04
Enforce pre-delegation authorization gates before agent dispatch.
SkillCFreshness: C★ 51Claude Code +9
Signals
- Stars
- 51
- Forks
- 16
- Open issues
- 21
- Last commit (this folder)
- May 7, 2026 (5 months ago)
From GitHub (fetched Oct 24, 2026).
Details
- Version
- 2.0.0
- License
- MIT
- Repository
- github.com/shynlee04/hivemind-plugin
- First seen
- Oct 24, 2026
- Last checked
- Oct 24, 2026
- Source
- GitHubgithub-graphql
More from shynlee04
- 0.578score
hm-l2-coordinating-loop
shynlee04
Coordinate multi-agent dispatch with validation gates and handoff protocols.
SkillCFreshness: C★ 51Claude Code +9 - 0.578score
hm-l3-integration-contracts
shynlee04
Formal bidirectional contract authority between skills and agents.
SkillCFreshness: C★ 51Claude Code +9 - 0.578score
hf-l2-skill-synthesis
shynlee04
Use when synthesizing skills from GitHub repos, classifying skill patterns, building template libraries, generating eval frameworks, or creating eval-driven…
SkillCFreshness: C★ 51Claude Code +9 - 0.578score
hf-l2-agent-composition
shynlee04
Teaches agents how to compose specialist agents from user intent using XML markup grammar, step protocols, and structured return formats.
SkillCFreshness: C★ 51Claude Code +9 - 0.578score
hm-l3-detective
shynlee04
Investigate codebases with SCAN, READ, and DEEP reading modes.
SkillCFreshness: C★ 51Claude Code +9 - 0.578score
hm-l3-synthesis
shynlee04
Compress research findings into actionable artifacts with tiered reduction.
SkillCFreshness: C★ 51Claude Code +9
Cite this page
gate-l3-lifecycle-integration skill. Agent Mart. https://agentmart.dev/skills/shynlee04-gate-l3-lifecycle-integration (last updated 2026-10-24)