ln-820-dependency-optimization-coordinator
Upgrades dependencies across all detected package managers. Use when updating npm, NuGet, or pip packages project-wide.
MCP get_skill({ skillId: "ln-820-dependency-optimization-coordinator-ac288485" })Use this skill with your agent
Create a free account and connect via MCP
> **Paths:** File paths (`references/`, `../ln-*`) are relative to this skill directory.
# ln-820-dependency-optimization-coordinator
**Type:** L2 Domain Coordinator
**Category:** 8XX Optimization
Runtime-backed coordinator for cross-stack dependency upgrades. Detects package managers, delegates to one worker per manager, records machine-readable worker summaries, and emits a final coordinator summary.
---
## Overview
| Aspect | Details |
|--------|---------|
| **Input** | Project path plus optional upgrade policy |
| **Output** | Aggregated dependency upgrade report with per-worker results |
| **Workers** | ln-821 (npm), ln-822 (nuget), ln-823 (pip) |
| **Runtime** | `.hex-skills/dependency/runtime/runs/{run_id}/` |
---
## Workflow
**Phases:** Pre-flight -> Detect Package Managers -> Security Audit -> Delegate Upgrades -> Collect Results -> Verify Summary -> Report
---
## Runtime Contract
**MANDATORY READ:** Load `references/ci_tool_detection.md`
**MANDATORY READ:** Load `references/coordinator_runtime_contract.md`, `references/dependency_runtime_contract.md`, `references/coordinator_summary_contract.md`
Runtime CLI:
```bash
node references/scripts/dependency-runtime/cli.mjs start --identifier repo-deps --manifest-file <file>
node references/scripts/dependency-runtime/cli.mjs status --identifier repo-deps
node references/scripts/dependency-runtime/cli.mjs checkpoint --phase PHASE_3_DELEGATE_UPGRADES --payload '{...}'
node references/scripts/dependency-runtime/cli.mjs record-worker-result --payload '{...}'
node references/scripts/dependency-runtime/cli.mjs record-summary --payload '{...}'
node references/scripts/dependency-runtime/cli.mjs advance --to PHASE_4_COLLECT_RESULTS
node references/scripts/dependency-runtime/cli.mjs complete
```
Required state fields:
- `worker_plan`
- `worker_results`
- `child_runs`
- `verification_passed`
- `report_ready`
- `summary_recorded`
Domain checkpoints:
- `PHASE_1_DETECT_PACKAGE_MANAGERS`: detected managers, indicator files, skipped managers
- `PHASE_2_SECURITY_AUDIT`: per-manager audit verdicts, blocking findings, release-age policy
- `PHASE_3_DELEGATE_UPGRADES`: one `child_run` per delegated worker with worker name, identifier, `runId`, and `summaryArtifactPath`
- `PHASE_4_COLLECT_RESULTS`: recorded worker summaries plus unresolved failures or warnings
- `PHASE_5_VERIFY_SUMMARY`: final report path, verification verdict, summary readiness
Guard rules:
- do not advance from `PHASE_3_DELEGATE_UPGRADES` until every planned worker emitted a valid `dependency-worker` summary
- do not complete until the final report checkpoint exists and the `dependency-coordinator` summary was recorded
- consume worker JSON summaries only; never infer worker status from prose output
---
## Phase 0: Pre-flight
Confirm the project is a valid candidate for dependency work before starting the runtime.
| Check | Method | Block if |
|-------|--------|----------|
| Manifest exists | Runtime `start` validation | Missing |
| Project path exists | File inspection | Missing |
| Upgrade policy provided | Manifest or defaults | No |
| Existing active run for identifier | Runtime active pointer | Conflicting active run |
Default options:
| Option | Default | Meaning |
|--------|---------|---------|
| `upgradeType` | `major` | major, minor, or patch |
| `allowBreaking` | `true` | allow major-version migrations |
| `minimumReleaseAge` | `14` | skip very recent releases unless security requires them |
| `testAfterUpgrade` | `true` | workers verify build/tests after changes |
---
## Phase 1: Detect Package Managers
Detect one worker target per package-manager family.
| Package Manager | Indicator Files | Worker |
|-----------------|-----------------|--------|
| npm | `package.json` + `package-lock.json` | ln-821 |
| yarn | `package.json` + `yarn.lock` | ln-821 |
| pnpm | `package.json` + `pnpm-lock.yaml` | ln-821 |
| nuget | `*.csproj` or `*.sln` | ln-822 |
| pip | `requirements.txt` | ln-823 |
| poetry | `pyproject.toml` + `poetry.lock` | ln-823 |
| pipenv | `Pipfile` + `Pipfile.lock` | ln-823 |
Checkpoint payload must include:
- `detected_managers`
- `indicator_paths`
- `worker_plan`
- `skipped_reasons`
---
## Phase 2: Security Audit
Perform lightweight pre-flight security and freshness checks before delegating heavy upgrade work.
| Manager Family | Command | Block Condition |
|----------------|---------|-----------------|
| Node.js | `npm audit --audit-level=high` or manager equivalent | Critical vulnerability with no allowed override |
| NuGet | `dotnet list package --vulnerable` | Critical vulnerability with no allowed override |
| Python | `pip-audit --json` or manager equivalent | Critical vulnerability with no allowed override |
Release-age gate:
| Option | Default | Description |
|--------|---------|-------------|
| `minimumReleaseAge` | `14 days` | Skip packages released too recently |
| `ignoreReleaseAge` | `false` | Override for urgent security patches |
Checkpoint payload must include:
- `audit_results`
- `blocking_findings`
- `release_age_policy`
- `managers_cleared_for_delegation`
---
## Phase 3: Delegate Upgrades
Delegate one child run per worker family. Child runs must be deterministic and artifact-driven.
Delegate using the concrete worker identities selected by the routing table below. Do not synthesize family placeholders or guessed skill IDs in prompts.
Delegation context:
| Field | Type | Description |
|-------|------|-------------|
| `projectPath` | string | Absolute path to target project |
| `packageManager` | enum | npm, yarn, pnpm, nuget, pip, poetry, pipenv |
| `identifier` | string | Stable worker identifier inside the run |
| `runId` | string | Deterministic child run id |
| `summaryArtifactPath` | string | Exact JSON path for the worker summary |
| `options` | object | Upgrade policy, verification flags, safety flags |
Worker selection:
| Manager Family | Worker | Notes |
|----------------|--------|-------|
| npm, yarn, pnpm | ln-821-npm-upgrader | One child run per detected Node manager |
| nuget | ln-822-nuget-upgrader | One child run for .NET |
| pip, poetry, pipenv | ln-823-pip-upgrader | One child run per detected Python manager |
After launching each worker:
1. Checkpoint `child_run` under `PHASE_3_DELEGATE_UPGRADES`.
2. Wait for the emitted `dependency-worker` summary envelope.
3. Record the worker summary with `record-worker-result`.
---
## Phase 4: Collect Results
Aggregate validated worker summaries only.
Worker summary fields consumed by the coordinator:
| Field | Description |
|-------|-------------|
| `producer_skill` | worker identity (`ln-821`, `ln-822`, `ln-823`) |
| `summary_kind` | must be `dependency-worker` |
| `identifier` | stable worker identifier |
| `payload.status` | completed, partial, or failed |
| `payload.upgrades` | applied upgrades with before/after versions |
| `payload.warnings` | non-blocking issues |
| `payload.verification` | build/test verification result |
| `payload.artifact_path` | worker-owned durable report path, if any |
Collection output:
- `worker_results`
- `success_count`
- `partial_count`
- `failed_count`
- `blocking_failures`
---
## Phase 5: Verify Summary
Prepare the final durable report and verify the coordinator can finish deterministically.
Verification checklist:
- every planned worker produced one valid summary envelope
- aggregate counts match recorded worker results
- final report path exists or is ready to be written
- `report_ready` and `verification_passed` are true before completion
Failure handling:
1. Keep successful worker results intact.
2. Mark failed workers explicitly in the coordinator report.
3. Do not invent rollback actions beyond what workers already verified.
---
## Phase 6: Report
Coordinator report schema:
| Field | Description |
|-------|-------------|
| `package_managers` | detected managers handled in this run |
| `workers_activated` | delegated workers |
| `total_packages` | packages analyzed across workers |
| `upgraded` | successful upgrades |
| `skipped` | already latest or policy-skipped packages |
| `failed` | packages or worker runs that failed |
| `breaking_changes` | major-version upgrades or migrations |
| `verification_passed` | aggregate verification verdict |
| `per_worker[]` | machine-readable worker result summaries |
| `warnings[]` | cross-worker warnings |
Completion sequence:
1. Write the durable report.
2. Checkpoint the report path and verification verdict.
3. Record the `dependency-coordinator` summary envelope with `record-summary`.
4. Complete runtime only after the report checkpoint and coordinator summary exist.
---
## Configuration
```yaml
Options:
upgradeType: major # major | minor | patch
allowBreaking: true
minimumReleaseAge: 14
auditLevel: high # none | low | moderate | high | critical
testAfterUpgrade: true
buildAfterUpgrade: true
rollbackOnFailure: true
skipDev: false
skipOptional: true
```
---
## Error Handling
Recoverable:
| Error | Recovery |
|-------|----------|
| Peer dependency conflict | Keep worker result as partial, continue collecting |
| Build failure in one worker | Preserve failure, continue other workers |
| Network timeout | Worker retries locally, then reports failure |
Fatal:
| Error | Action |
|-------|--------|
| No package managers found | Finish with empty-result report |
| Runtime validation failure | Pause run and require intervention |
| Missing worker summary for planned child run | Do not advance from collection |
---
## References
- [breaking_changes_patterns.md](references/breaking_changes_patterns.md)
- [security_audit_guide.md](references/security_audit_guide.md)
---
**TodoWrite format (mandatory):**
```text
- Detect package managers (in_progress)
- Delegate ln-821-npm-upgrader child runs (pending)
- Delegate ln-822-nuget-upgrader child runs (pending)
- Delegate ln-823-pip-upgrader child runs (pending)
- Aggregate dependency-worker summaries (pending)
```
## Worker Invocation (MANDATORY)
**Host Skill Invocation:** `Skill(skill: "...", args: "...")` is mandatory delegation.
- Claude: call the Skill tool exactly as shown.
- Codex: if no Skill tool exists, locate the named skill in available skills, read its `SKILL.md`, treat `args` as `$ARGUMENTS`, execute that skill workflow, then return here with its result/artifact.
- Do not inline worker logic or mark the worker complete without executing the target skill.
| Phase | Worker | Context |
|-------|--------|---------|
| 3 | ln-821-npm-upgrader | Isolated child run with `packageManager`, `runId`, and exact `summaryArtifactPath` |
| 3 | ln-822-nuget-upgrader | Isolated child run with `packageManager`, `runId`, and exact `summaryArtifactPath` |
| 3 | ln-823-pip-upgrader | Isolated child run with `packageManager`, `runId`, and exact `summaryArtifactPath` |
All workers: start the child runtime, checkpoint the `child_run` metadata, then invoke the worker skill explicitly and consume the emitted `dependency-worker` summary envelope via `record-worker-result`.
```text
# One invocation per detected package manager (sequential per family):
node references/scripts/dependency-runtime/cli.mjs start --skill {worker} --identifier {packageManager} --manifest-file {workerManifestPath} --run-id {childRunId} --summary-artifact-path {childSummaryArtifactPath}
node references/scripts/optimization-runtime/cli.mjs checkpoint --phase PHASE_3_DELEGATE --payload '{"child_run":{"worker":"{worker}","run_id":"{childRunId}","summary_artifact_path":"{childSummaryArtifactPath}","package_manager":"{packageManager}"}}'
Skill(skill: "{worker}", args: "{packageManager} --run-id {childRunId} --summary-artifact-path {childSummaryArtifactPath}")
Read {childSummaryArtifactPath}
node references/scripts/optimization-runtime/cli.mjs record-worker-result --payload-file {childSummaryArtifactPath}
```
Worker token substitution: `{worker}` is one of `ln-821-npm-upgrader`, `ln-822-nuget-upgrader`, `ln-823-pip-upgrader`.
---
## Definition of Done
- [ ] Runtime started with a validated manifest and stable identifier
- [ ] Package managers detected from project indicators
- [ ] Pre-flight security and release-age checks completed
- [ ] One child run delegated per planned worker family
- [ ] Every child run emitted a valid `dependency-worker` summary
- [ ] Coordinator report aggregates per-worker upgrades, warnings, and verification results
- [ ] Final `dependency-coordinator` summary recorded before completion
---
## Phase 7: Meta-Analysis
Optional reference: load `references/meta_analysis_protocol.md` only when the user asks for post-run meta-analysis or protocol-formatted run reflection.
Skill type: `optimization-coordinator`. When requested, run after all phases complete. Output to chat using the `optimization-coordinator` format.
---
**Version:** 1.1.0
**Last Updated:** 2026-01-10Related Skills
More skills in Software Engineering
Accessibility Standards
Comprehensive web accessibility standards based on WCAG 2.2 AA, with 38+ anti-patterns, legal enforcement context (EAA, ADA Title II), WAI-ARIA patterns, and framework-specific fixes for modern web frameworks and libraries.
Accord
Authoring unified specification packages across Business/Development/Design teams via staged elaboration (L0 Vision → L1 Requirements → L2 Team Detail → L3 Acceptance Criteria). No code. Use when authoring cross-team specs, building L0-L3 packages, or aligning Biz/Dev/Design on a single source of truth.
Acquire Codebase Knowledge
Use this skill when the user explicitly asks to map, document, or onboard into an existing codebase. Trigger for prompts like "map this codebase", "document this architecture", "onboard me to this repo", or "create codebase docs". Do not trigger for routine feature implementation, bug fixes, or narrow code edits unless the user asks for repository-level discovery.
Acreadiness Assess
Run the AgentRC readiness assessment on the current repository and produce a static HTML dashboard at reports/index.html. Wraps `npx github:microsoft/agentrc readiness` and hands off rendering to the @ai-readiness-reporter custom agent. Supports policies (--policy) for org-specific scoring. Use when asked to assess, audit, or score the AI readiness of a repo.
Acreadiness Generate Instructions
Generate tailored AI agent instruction files via AgentRC instructions command. Produces .github/copilot-instructions.md (default, recommended for Copilot in VS Code) plus optional per-area .instructions.md files with applyTo globs for monorepos. Use after running /acreadiness-assess to close gaps in the AI Tooling pillar.
Acreadiness Policy
Help the user pick, write, or apply an AgentRC policy. Policies customise readiness scoring by disabling irrelevant checks, overriding impact/level, setting pass-rate thresholds, or chaining org baselines with team overrides. Use when the user asks about strict mode, AI-only scoring, custom weights, CI gating, or wants org-wide standardisation.
Explore Other Categories
Skills from other categories with shared topics
Pipeline Orchestrator
Drives a Story through full pipeline (tasks, validation, execution, quality). Use when executing a Story end-to-end from kanban board.
Story Coordinator
Creates, replans, or appends 5-10 Stories per Epic with standards research and multi-epic routing. Use when Epic needs Story decomposition.
Story Replanner
Replans Stories by comparing IDEAL vs existing (KEEP/UPDATE/OBSOLETE/CREATE). Use when Epic requirements changed and Stories need realignment.