esonhugh/marketplace/close-case
Claude Code and agent skill by esonhugh/marketplace
- community
Checking that the SKILL.md file is still in the repository.
The assistant should use this when the user asks to close, finalize, force-close, wrap up with a final conclusion, or produce final Detective case findings. It closes only after proof readiness or explicit force approval. Do not use for status-only review (review-board), further investigation (investigate), blocker discussion (discuss-case), new case creation (open-case), or pre-case scoping (brainstorm).
Install
npx vskill@latest install esonhugh/marketplace/close-caseSKILL.md
Close Case
Close a Detective MCP-backed case only through MCP tools. Follow ../references/detective-core-protocol.md for proof completion semantics, canonical state, StopHook deactivation, handoffs, and export policy.
<HARD_GATE name="closure-authorization"> Normal closure requires detective_completion_gate approval and a final summary naming conclusion, confidence, key evidence chain, completed coverage, and unresolved caveats. If the gate is blocked, do not close; present blockers and offer continue, discuss, or force-close options. </HARD_GATE>
<HARD_GATE name="force-close-approval"> force=true requires explicit user approval for a partial or forced close in the current conversation. Include the force reason and unresolved caveats in the summary; never infer force approval from frustration, silence, or a generic request to wrap up. </HARD_GATE>
<IMPORTANT name="proof-lifecycle-alignment"> Before normal closure, verify exactly one confirmed hypothesis, at least one confirmed evidence node with a direct supports edge to it, questions and alternatives resolved/rejected/stale, actions done/cancelled, and coverage complete. Use detective_completion_gate as the readiness authority; use detective_evaluate_proof only when preserving a durable final proof snapshot. </IMPORTANT>
<IMPORTANT name="goal-completion-alignment"> When a SetGoal or equivalent goal tool is available, treat detective_completion_gate approval as the goal-completion gate. Session goals are orchestration hints and do not replace MCP closure state. </IMPORTANT>
<HARD_GATE name="stop-fallback-deactivation-on-close"> When closure is completed, blocked waiting for user force-close approval, or paused by user request, emit <DETECTIVE-STOP-FALLBACK status="inactive" case-id="<case_id>" reason="<closed|blocked|pause>">. This only disengages the transcript-activated fallback and must not be written to .detective/ files. </HARD_GATE>
Required Inputs
Obtain or draft case_id, summary, approved_by (user role/name if provided, else assistant after gate approval), and force=false unless explicitly approved.
Exact Workflow
- Read final state:
detective_case_status(case_id="<case id>")detective_graph_overview(case_id="<case id>")detective_coverage_status(case_id="<case id>")detective_completion_gate(case_id="<case id>")
- If a durable final proof snapshot is requested or needed for final reporting, call:
detective_evaluate_proof(case_id="<case id>", summary="<draft final summary>")
- If the completion gate is blocked, present blockers and route:
- proof/coverage/action gaps → offer
/detective:investigate; - missing user decision/scope/caveat approval → offer
/detective:discuss-case; - explicit partial-finalization request → ask for force-close approval.
- For normal closure:
detective_close_case(case_id="<case id>", summary="<final summary>", approved_by="<approver>", force=false)
- For explicit partial/forced closure:
detective_close_case(case_id="<case id>", summary="<summary including caveats and force reason>", approved_by="<approver>", force=true)
- After successful close, export:
detective_export_markdown(case_id="<case id>")detective_export_mermaid(case_id="<case id>", diagram="full", focus_node_id=null)- Optional focused diagram only if useful:
detective_export_mermaid(case_id="<case id>", diagram="hypothesis-chain", focus_node_id="<node id>")
- Emit inactive marker with reason
closed,blocked, orpauseas applicable:
<DETECTIVE-STOP-FALLBACK status="inactive" case-id="<case_id>" reason="closed"><DETECTIVE-STOP-FALLBACK status="inactive" case-id="<case_id>" reason="blocked">
Final Summary Template
Response Format
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 repositoryCheckingplugins/detective-plugin/skills/close-case/SKILL.md
Popularity
- 7-day trend score
- 0
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/esonhugh/marketplace/close-case)