maxwell2732/claudecode-research-harness-workflow/session-init
Claude Code and agent skill by maxwell2732/claudecode-research-harness-workflow
- community
Checking that the SKILL.md file is still in the repository.
Internal sub-skill for session startup checks, Plans.md status, git state, and harness-mem resume pack. Invoked by session/startup workflows only. Do NOT load for: implementation, reviews, or mid-session tasks.
Install
npx vskill@latest install maxwell2732/claudecode-research-harness-workflow/session-initSKILL.md
Session Init Skill
A skill for checking the environment and understanding the current task status at session start.
Invocation Conditions
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:
- "Start session"
- "Begin work"
- "Start today's work"
- "Check status"
- "What should I do?"
- "start session"
- "what should I work on?"
Overview
The Session Init skill automatically checks the following at Claude Code session start:
- Git state: Current branch, uncommitted changes
- Plans.md: In-progress tasks, requested tasks
- AGENTS.md: Role assignments, review of prohibitions
- Previous session: Review of carry-over items
- Latest snapshot: Summary of progress snapshot and diff from previous session
Execution Steps
Step 0: File State Check (Automatic Cleanup)
Check file sizes before session start:
Display a suggestion if cleanup is needed (does not affect work).
Step 0.5: Legacy Local Memory Compatibility Handling (Optional)
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.
Step 0.7: Unified Harness Memory Resume Pack (Required)
Retrieve resume context from the Codex / Claude / OpenCode shared DB (~/.harness-mem/harness-mem.db).
Required call:
Operational rules:
- Always specify current project name for
project - Obtain
session_idin order:$CLAUDE_SESSION_ID→.session_idin.claude/state/session.json - Use of the first result from
harness_mem_sessions_list(project, limit=1)is limited to read-only (resume check); do not use for writes viarecord_checkpoint/finalize_session - Inject retrieval results into session start context
- On retrieval failure, check daemon state with
harness_mem_health(), note the failure, and continue - Recovery order:
scripts/harness-memd doctor→scripts/harness-memd cleanup-stale→scripts/harness-memd start
Step 1: Environment Check
Run the following in parallel:
Step 2: Understand Task Status
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
Step 3: Output Status Report
Output Format
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 |
Related Commands
/work- Task execution (parallel execution supported)/sync-status- Plans.md progress summary/maintenance- Automatic file cleanup
Notes
- Always check AGENTS.md: Understand role assignments before starting work
- If Plans.md is missing: Guide to
/harness-init - If previous work was interrupted: Confirm whether to continue
Security
- Security scanPassed, 90 of 100Pattern checks on the skill's files.
- Model reviewPassed
Quality evaluation
- Task evaluationNot runNo evaluation yet. Run one in Skill Studio.
Source
- SKILL.md in the repositoryCheckingskills/session-init/SKILL.md
Popularity
- Repository stars
- 40
- Repository forks
- 49
- 7-day trend score
- 0
Stars and forks are for the whole GitHub repository, not this skill alone. Not refreshed yet.
Works with
Any agent vskill installs to: 56 targets in all. These 39 have a page on this site.
- Claude Code
- Cursor
- GitHub Copilot
- Windsurf
- Codex
- Gemini CLI
- Cline
- Amp
- Roo Code
- Goose
- Continue
- Kilo Code
- Junie
- Trae
- OpenHands
- Qwen Code
- AI Agent
- Tabnine Chat
- Sourcegraph Cody
- Amazon Q
- Firebase Genkit
- Sweep
- Devin
- Phind
- Mentat
- Claude Desktop
- Aider
- GPT Pilot
- Plandex
- bolt.new
- v0
- Replit Agent
- Anse
- TypingMind
- BetterChatGPT
- LibreChat
- Warp AI
- Pieces
- Fig AI
Embed badge
[](https://verified-skill.com/skills/maxwell2732/claudecode-research-harness-workflow/session-init)