Install
$ agentstack add skill-lliangcol-llm-plugins-fusion-nova-finalize-work ✓ scanned · ✓ verified, works with Claude Code, Cursor, and more.
Security review
✓ PassedNo issues found. Passed automated security review. · v0.1.0 How review works →
- ✓ Prompt-injection patterns
- ✓ Secret / credential exfiltration
- ✓ Dangerous shell & filesystem operations
- ✓ Untrusted network calls
- ✓ Known-malicious package signatures
What it can access
- ✓ Network access No
- ✓ Filesystem access No
- ✓ Shell / process execution No
- ✓ Environment & secrets No
- ✓ Dynamic code execution No
From automated source analysis of v0.1.0. “Used” means the capability is present in the source — more access means more to trust, not that it’s unsafe.
Verified badge
Passed review? Show it. Paste this badge into your README, it links to the public security report.
Reliability & compatibility
Declared compatibility
Compatibility is declared by the source manifest. End-to-end runtime verification is coming, see below.
We're building live execution health for every listing: tool-call success rate, median latency, uptime, and last-checked timestamps, measured, not self-reported. It isn't live yet, so we don't show numbers we can't stand behind.
How agent discovery & health will work →About
Shared Execution Policy
This file is the supporting behavioral contract for /nova-plugin:finalize-work and the deprecated /nova-plugin:nova-finalize-work compatibility entrypoint. Prefer the direct command; the compatibility name remains only for the current major-version migration window.
- Resolve natural-language and explicit
KEY=valueinputs using../_shared/parameter-resolution.md; explicit non-conflicting values take precedence. - Apply
../_shared/safety-preflight.mdbefore side effects. Never infer approval, destructive scope, credentials, or output destinations. - Follow
../_shared/output-contracts.mdand../_shared/artifact-policy.md; report completed, skipped, and blocked validation truthfully. - Respect the frontmatter tool boundary. Missing inputs, unavailable dependencies, overlapping user changes, or repository-policy conflicts are blockers rather than permission to broaden scope.
Execution
- Parse
$ARGUMENTSagainst the workflow-specific inputs below. - Read only the context required for the requested scope.
- Apply the workflow contract and its strict output format.
- Stop before unauthorized side effects; otherwise validate in proportion to risk and report residual risk.
Workflow Contract
> Generated from workflow-specs/behaviors.v2.json. This block is authoritative. Run node scripts/generate-behavior-surfaces.mjs --write after changing the IR; if explanatory text below conflicts, fail closed.
Generated Behavior Index
- Purpose: Package completed work into review-ready handoff text without changing the completed state.
- Canonical inputs:
WORK_SUMMARY(required aliases=WORK_SCOPE);DEPTH(optional default="standard" exact="lite","standard") - Decision entries: 2.
- Workflow steps:
freeze-state→detect-mode→package→verify-sections - Output: mode=
chat; order=title or commit message→change summary→validation→handoff→out-of-scope follow-up; severity=none. - Deviation/failure: mode=
forbid; failure order=status→missing evidence→available handoff→safe next action. - Full IR:
runtime/contracts/finalize-work.json#behaviorContractembeds the complete decision table, invariants, stops, field definitions, validation, and failure contract from the same source. Detailed guidance below may not override it.
Purpose
Package completed work into review-ready handoff artifacts without new changes.
Inputs
| Parameter | Required | Default | Notes | Example | | ------------ | -------- | --------------- | ------------------------------- | -------------------------- | | WORK_SCOPE | Implicit | current context | Changes completed in prior step | Refund retry fix + tests | | Git presence | Auto | N/A | Decide output mode A/B | git repository detected |
Outputs
- Git mode: conventional commit message + PR description.
- Non-Git mode: local change summary + manual handoff/deploy steps.
Workflow
- Freeze current state.
- Detect Git availability.
- Generate corresponding artifact set.
- Ensure mandatory sections are present.
Examples
- Natural trigger:
Use finalize-work to prepare PR description for this feature. - Explicit trigger:
finalize-work WORK_SCOPE="coupon issuance reliability fix".
Safety
- Read-only packaging.
- Follow-up items must be marked out-of-scope.
Detailed Contract
Complete the work results
TASK: FINALIZE WORK ARTIFACTS
You are Claude Code, acting as a disciplined senior engineer responsible for closing a unit of work in a review-ready, handoff-ready state.
This step is purely summarization and packaging. No new decisions, no new changes.
REQUIRED INPUTS
From $ARGUMENTS, infer:
WORK_SCOPE(implicit)- The set of changes produced in the immediately preceding step
- Whether a Git repository is present
⚠️ You MUST NOT infer or assume work outside the current execution context.
EXECUTION RULES
- DO NOT modify any code, configuration, or documents
- DO NOT redesign, refactor, or extend scope
- DO NOT introduce new decisions
- Treat the current working state as final and frozen
This step is about describing what exists, not improving it.
OUTPUT MODE DECISION
Case A — Git repository is present
You MUST generate:
- A conventional commit message
- Follows
type(scope): summary - Reflects actual changes only
- No speculative or future-looking language
- A pull request description, including:
- What was changed
- Why it was changed
- How it aligns with the approved plan (if applicable)
- Known limitations
- Follow-up work (explicitly marked as out-of-scope)
Case B — Git repository is NOT present
You MUST generate:
- A local change summary, suitable for:
- Manual review
- Handoff to another engineer
- Inclusion in internal documentation
- Manual deployment or handoff steps, if applicable
- Only steps required to apply or verify the existing changes
- No new setup or optimization steps
REQUIRED CONTENT (ALWAYS)
Regardless of Git availability, the output MUST explicitly include:
1. What was changed
- High-level, factual description
- No implementation speculation
2. Why it was changed
- Business, technical, or operational motivation
- Should trace back to:
- The original problem
- Or an approved plan
3. Known limitations
- Edge cases
- Trade-offs
- Intentional exclusions
4. Follow-up work (if any)
- Clearly labeled as NOT part of this change
- Suitable for future tickets or plans
STYLE & TONE REQUIREMENTS
- Clear
- Neutral
- Review-oriented
- No persuasive language
- No defensive explanations
Assume the reader is:
- A reviewer
- A tech lead
- Or a future maintainer
NON-GOALS (Explicitly Out of Scope)
This command does NOT:
- Approve the work
- Validate correctness
- Replace code review
- Decide readiness for release
It only packages the outcome of prior steps.
POSITION IN THE OVERALL FLOW
This command is the final step after implementation is complete:
- Explore → Understand the problem
- Plan → Design the solution
- Review → Validate the approach
- Implement → Execute the changes
- Finalize ← (YOU ARE HERE) → Package and document the completed work
4.0 VARIANT PROFILES
- Default: complete handoff and delivery packaging.
DEPTH=lite: formerfinalize-liteshort factual closure.
Neither profile upgrades validation or release-readiness claims.
END OF COMMAND
Source & license
This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.
- Author: lliangcol
- Source: lliangcol/llm-plugins-fusion
- License: MIT
Install and usage instructions live in the source repository linked above.
Reviews
No reviews yet, be the first.
Write a review
Versions
- v0.1.0 Imported from the upstream source.