Skills DirectorySkills Directory
SkillsLearnSecurityCategoriesDocsCommunityBlog
Sign InSubmit Skill
Skills Directory

Security-tested agent skills for Claude, coding agents, and AI workflows.

Directory

  • Browse Skills
  • All Skills A–Z
  • Claude Skills
  • Claude Code Skills
  • Agent Skills
  • Categories
  • Submit a Skill

Learn

  • Learn Hub
  • Install Claude Skills
  • Write SKILL.md
  • Skills vs MCP
  • Directories Compared

Security

  • Security
  • Methodology
  • Secure Claude Skills
  • Security Badges

Company

  • About
  • Community
  • Blog
  • API Docs
  • Advertise

2026 Skills Directory. All rights reserved.

Back to skills

Close

ASecurity

Use when verified work is ready for QA, acceptance, checkpointing, learning, recommendation, and handoff.

22 stars
0 votes
0 copies
0 views
Added 9/20/2026
testinggogit

Works with

cursorcli

Security Analysis

A100/100

Scanned 9/20/2026

Install to Claude Code

$npx -y skills add lev-os/agents --skill close --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Close?

Add the live security badge to your README — it updates automatically with every re-scan.

Security grade badge for Close
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/lev-os-close/badge)](https://www.skillsdirectory.com/skills/lev-os-close)

More formats (shields.io, HTML) on the badges page.

Download Zip
Files
SKILL.md
---
name: close
description: Use when verified work is ready for QA, acceptance, checkpointing, learning, recommendation, and handoff.
---

# /close - Close Lane

Close the active entity: capture leftovers, verify acceptance, checkpoint, learn,
recommend, and hand off.

## Entity Reconciliation

After authorized material progress, reconcile touched and causally affected
artifacts before routing, handoff, or final response. Track entity ref,
island/provider locator (a path for file storage), basis/evidence, and the
reason/action due. Update through the verified owning CLI/adapter; use a
write-authorized skill fallback only when that operation is unavailable.
Record updated, no_change(reason), or blocked(reason), preserving unresolved
refs. Reading or mentioning a path alone creates no update obligation.
Read-only work reports pending changes only. Reminders grant no write authority;
task status stays with the bound tracker. Update only artifacts whose content
or evidence changed; do not rewrite every referenced document.

## Work Link

Lifecycle lane: Close
Entity movement: `verified -> closed | monitoring | follow_up | blocked`
Workstream: update active workstream before reporting done
Upstream: `/exec`
Downstream: `/handoff`, `/work`
Router: `/work`
HUD: end with `🧬 {ws} ⚡{exec_count} 📥{capture_count} ⏸️{paused_count} ✅{done_count} | 🚦{gate}={score} | ⏭️ {next} | 🔁{loop_state}`

Apply acceptance against the active domain's plan or execution contract; task
DNA and software-specific proof gates apply to the SDLC overlay. Mark genuinely
inapplicable software gates N/A for non-coding work. Checkpoint Git only when
the deliverable is Git-backed and commit/push are within the authorized scope.

## Next Routes

Use the result to select the next owner. Show applicable alternatives in a
numbered `skill://<name>` table only when a decision remains; suggestions never
grant new execution, publishing, or checkpoint authority.

| Result | Next owner |
|---|---|
| Uncaptured intent or leftovers | `skill://capture` |
| Acceptance failed within the existing authorized contract | `skill://exec` using the same domain method |
| Coding contract must change | `skill://propose`; `skill://lev-plan` if the broader plan changes |
| Non-coding scope or plan must change | `skill://lev-plan` or `skill://interview`; proposal optional |
| Accepted or paused, context needs preservation | `skill://handoff` |
| Several completed efforts warrant synthesis | `skill://lev-learn` |
| Another effort or uncertain next owner | `skill://lev` |

## Protocol

```yaml
steps:
  - id: capture_gate
    action: Run capture in gate mode before closing.
    validation: "in_memory_count == 0."
    on_failure: "Route to /capture and stop close."

  - id: qa_acceptance
    action: Check output against the active domain plan or execution contract, acceptance criteria, declared verifier evidence, and applicable proof-gate verdicts.
    validation: "Every acceptance item and applicable proof gate is pass, fail, blocked, or explicitly not-tested with residual risk."
    on_failure: "Route failed criteria back to /exec."

  - id: proof_gate_verdict
    action: Seal the proof gates required by the active domain contract before acceptance.
    validation: "Applicable SDLC Pentagon, UltraQA, cleanup, and ai-slop-cleaner results or domain-specific equivalents are recorded; inapplicable gates have an N/A rationale, and residual risks are explicit."
    on_failure: "Do not accept; route to /exec or /propose depending on whether runtime evidence or proof contract is missing."

  - id: accept
    action: Decide whether the work is accepted.
    validation: "Confidence >= 0.85 for auto-accept, otherwise human approval is recorded."
    on_failure: "Do not commit or handoff as closed."

  - id: checkpoint
    action: For authorized Git-backed checkpoints, stage, commit, pull without rebase, and push when sealing verified work; otherwise record not_applicable or the authority boundary.
    validation: "Authorized checkpoint succeeds, is not_applicable, or its blocker/authority boundary is recorded."
    on_failure: "Keep entity blocked and do not claim closed."

  - id: learn
    action: Extract typed learnings with claim, evidence, applies_to, and propagates.
    validation: "At least one learning or explicit 'no new learning' is recorded."
    on_failure: "Add the learning block before handoff."

  - id: recommend
    action: Recommend keep, monitor, execute, or reject for related entities.
    validation: "Next lifecycle verb is clear."
    on_failure: "Route to /work for next-verb selection."

  - id: handoff
    action: Invoke /handoff to update workstream state and resume prompt.
    validation: "Workstream YAML reflects close result."
    on_failure: "Do not end without durable state."
```

## Trigger Record

When the workstream has a trigger stream, write:

```yaml
schema: lifecycle_trigger.v1
verb: close
entity_ref: lev://entity/work/workstream/<ws-id> | lev://entity/work/task/<task-id>
outputs:
  items: []
  learnings: []
  gates: []
  proof_gates:
    pentagon: pass | fail | blocked | not_applicable
    ultraqa: pass | fail | blocked | not_applicable
    ai_slop_cleaner: pass | fail | blocked | not_applicable
  artifacts: []
metrics:
  drift: <number>
  in_memory_count: 0
cursor:
  last_verb: close
  next_verb: null
recommend: []
```

## KPI Gate

| KPI | Move on |
|---|---|
| drift_score | `< 0.3` |
| hygiene_pass_rate | `>= 0.95` |
| learning_extracted | `>= 1` or explicit none |
| proof_gate_verdicts | all applicable gates pass or have explicit blocked/follow-up routing |

Max rework: 2. Three consecutive gate failures becomes blocked.

## Red Flags

- "Ready to push when you are."
- "Close can skip capture because the task is done."
- "A passing command means acceptance."
- "Repo-wide audit passed, so feature-local Pentagon proof is done."
- "Temporary QA harness cleanup can be handled later."
- "I'll write handoff later."
- "Pull with rebase."

## Related

- `/exec` provides verified work.
- `/capture` clears in-memory leftovers.
- `/handoff` updates workstream continuity.
- `/work` routes the next entity.

Attribution

lev-oslev-os
View sourceMore from lev-os →
SSkills DirectorySkills Directory

Your tool, in front of Claude Code builders.

3 founder slots · $299/mo · GSC-verified traffic · sponsors can never buy grades.

See placements

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments (0)

No comments yet. Be the first to comment!

SSkills DirectorySkills Directory

Your tool, in front of Claude Code builders.

3 founder slots · $299/mo · GSC-verified traffic · sponsors can never buy grades.

See placements

Related Skills

Screen Reader Testing

Practical guide to testing web applications with screen readers for comprehensive accessibility validation.

393431 votes

Python Testing

使用pytest、TDD方法、夹具、模拟、参数化和覆盖率要求的Python测试策略。

2456590 votes

Tdd Workflow

在编写新功能、修复错误或重构代码时使用此技能。强制执行测试驱动开发,包含单元测试、集成测试和端到端测试,覆盖率超过80%。

2456590 votes

Springboot Tdd

使用JUnit 5、Mockito、MockMvc、Testcontainers和JaCoCo进行Spring Boot的测试驱动开发。适用于添加功能、修复错误或重构时。

2456590 votes

Eval Harness

克劳德代码会话的正式评估框架,实施评估驱动开发(EDD)原则

2456590 votes
View all in testing →