Template

Files

lmiranda 0e70156e26 feat(projman): add plan-then-batch skill optimization

Separate cognitive work from mechanical API execution to reduce
skill-related token consumption by ~76-83% during sprint workflows.

Changes:
- Add batch-execution.md skill with 4-phase protocol
- Promote mcp-tools-reference and batch-execution to frontmatter
  for planner and orchestrator agents (auto-injected, zero re-read)
- Replace "Skills to Load" with phase-based "Skill Loading Protocol"
- Restructure planning-workflow.md Steps 8-10 for batch execution
- Update agent matrix in CLAUDE.md and docs/CONFIGURATION.md
- Add Phase-Based Skill Loading documentation section
- Clean up .gitignore (transient files, dev symlinks)

Token impact:
- 6-issue sprint planning: ~76% reduction
- 10-issue sprint planning: ~80% reduction
- 8-issue status updates: ~83% reduction

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>

2026-02-03 19:57:10 -05:00

2.5 KiB

Raw Blame History

description, agent

description	agent
Begin sprint execution with relevant lessons learned from previous sprints	orchestrator

Start Sprint Execution

Skills Required

skills/mcp-tools-reference.md (frontmatter — auto-injected)
skills/batch-execution.md (frontmatter — auto-injected)
skills/branch-security.md
skills/sprint-approval.md
skills/dependency-management.md
skills/lessons-learned.md
skills/git-workflow.md
skills/progress-tracking.md
skills/runaway-detection.md
skills/sprint-lifecycle.md

Purpose

Initiate sprint execution. The orchestrator agent verifies approval, analyzes dependencies for parallel execution, searches relevant lessons, and coordinates task dispatch.

Invocation

Run /sprint-start when ready to begin executing a planned sprint.

Flags:

--force — Bypass approval gate (emergency only, logged to milestone)

Workflow

Execute the sprint start workflow:

Verify Sprint Approval & Lifecycle State (required) - Check milestone for approval record. STOP if missing unless --force flag provided. Also verify lifecycle state is Sprint/Planning per skills/sprint-lifecycle.md. Set Sprint/Executing after verification passes.
Detect Checkpoints - Check for resume points from interrupted sessions
Fetch Sprint Issues - Get open issues from milestone
Analyze Dependencies - Use get_execution_order for parallel batches
Search Relevant Lessons - Find applicable past experiences
Dispatch Tasks - Parallel when safe, sequential when file conflicts exist

File Conflict Prevention: Before parallel dispatch, check target files for overlap. Sequentialize tasks that modify the same files.

Branch Isolation: Each task runs on its own branch (feat/<issue>-<desc>).

Sequential Merge: After completion, merge branches sequentially to detect conflicts.

Visual Output

╔══════════════════════════════════════════════════════════════════╗
║  📋 PROJMAN                                                      ║
║  ⚡ EXECUTION                                                    ║
║  [Sprint Name]                                                   ║
╚══════════════════════════════════════════════════════════════════╝

2.5 KiB Raw Blame History