From claudecode-research-harness-workflow
Internal session startup skill that checks git state, Plans.md status, and harness-mem resume pack. Not user-invocable; triggered by session/startup workflows only.
How this skill is triggered — by the user, by Claude, or both
Slash command
/claudecode-research-harness-workflow:session-initThis skill is limited to the following tools:
The summary Claude sees in its skill listing — used to decide when to auto-load this skill
A skill for checking the environment and understanding the current task status at session start.
A skill for checking the environment and understanding the current task status at session start.
This skill is called internally from session / SessionStart-type flows.
The user-facing entry point is /session or the normal session start flow.
Legacy trigger phrases:
The Session Init skill automatically checks the following at Claude Code session start:
Check file sizes before session start:
# Check Plans.md line count
if [ -f "Plans.md" ]; then
lines=$(wc -l < Plans.md)
if [ "$lines" -gt 200 ]; then
echo "⚠️ Plans.md has ${lines} lines. Recommend organizing with 'organize it'"
fi
fi
# Check session-log.md line count
if [ -f ".claude/memory/session-log.md" ]; then
lines=$(wc -l < .claude/memory/session-log.md)
if [ "$lines" -gt 500 ]; then
echo "⚠️ session-log.md has ${lines} lines. Recommend organizing with 'organize session log'"
fi
fi
Display a suggestion if cleanup is needed (does not affect work).
The current standard is the Unified Harness Memory in Step 0.7. Checking legacy local memory compatibility is generally unnecessary; reference it individually only for special migration checks.
Note: In normal operation, skip this step and treat the common DB Resume Pack as the sole resume entry point.
Retrieve resume context from the Codex / Claude / OpenCode shared DB (~/.harness-mem/harness-mem.db).
Required call:
harness_mem_resume_pack(project, session_id?, limit=5, include_private=false)
Operational rules:
projectsession_id in order: $CLAUDE_SESSION_ID → .session_id in .claude/state/session.jsonharness_mem_sessions_list(project, limit=1) is limited to read-only (resume check); do not use for writes via record_checkpoint / finalize_sessionharness_mem_health(), note the failure, and continuescripts/harness-memd doctor → scripts/harness-memd cleanup-stale → scripts/harness-memd startRun the following in parallel:
# Git state
git status -sb
git log --oneline -3
# Plans.md
cat Plans.md 2>/dev/null || echo "Plans.md not found"
# AGENTS.md key points
head -50 AGENTS.md 2>/dev/null || echo "AGENTS.md not found"
Extract from Plans.md:
cc:WIP - Tasks continuing from previous sessionpm:requested - Tasks newly requested by PM (compatible: cursor:requested)cc:TODO - Unstarted but assigned tasks## 🚀 Session Started
**Date/Time**: {{YYYY-MM-DD HH:MM}}
**Branch**: {{branch}}
**Session ID**: ${CLAUDE_SESSION_ID}
---
### 📋 Today's Tasks
**Priority tasks**:
- {{pm:requested (compatible: cursor:requested) or cc:WIP tasks}}
**Other tasks**:
- {{List of cc:TODO tasks}}
---
### ⚠️ Notes
{{Key constraints/prohibitions from AGENTS.md}}
---
**Ready to start work?**
At session start, present the following information concisely:
| Item | Content |
|---|---|
| Current branch | e.g., staging |
| Priority tasks | Most important 1–2 items |
| Notes | Summary of prohibitions |
| Next action | Specific suggestion |
/work - Task execution (parallel execution supported)/sync-status - Plans.md progress summary/maintenance - Automatic file cleanup/harness-initnpx claudepluginhub maxwell2732/claudecode-research-harness-workflow --plugin claudecode-research-harness-workflowInternal skill that checks git state, Plans.md status, AGENTS.md roles, and retrieves resume context from Unified Harness Memory (harness-mem DB) at session start. Reports oversized files and suggests cleanup.
Initializes Claude Code sessions by migrating legacy harness files to .claude-harness/, creating missing state files like memory rules and config, and preparing GitHub sync.
Creates bite-sized, testable implementation plans from specs or requirements, with file structure and task decomposition. Activates before coding multi-step tasks.