implementation-executor
Purpose
Execute an approved implementation plan safely and incrementally, with continuous verification and closeout updates.
Scope
This skill is for implementation and delivery.
- In scope: code changes, tests, validation, plan tracking, closeout documentation updates.
- Out of scope: creating a brand-new architecture direction without a plan, uncontrolled large rewrites, silent scope expansion.
Required Inputs
- Execution handoff package from
requirements-to-plan - Approved scope and constraints
- Acceptance criteria
- Validation checklist
- Rollback strategy
If the handoff package is missing critical sections, pause and request plan completion first.
Execution Strategy
Execute in small batches and validate after each batch.
- Read handoff package and restate scope boundaries
- Group tasks into minimum safe batches
- Execute one batch at a time
- Validate batch outcomes immediately
- Record deviations and rationale
- Continue only if batch exit criteria pass
- Perform final closeout and docs updates
Batch Rules
Each batch must include:
- Target files/modules
- Intended behavior change
- Risks and mitigations
- Validation commands/checks
- Expected outcome
After each batch, produce:
- What changed
- What was validated
- Pass/fail status
- Any deviation from plan
- Next action
Deviation Policy
If execution diverges from plan:
- Label as
minorormajor - Explain root cause
- Provide corrected path
- For major deviations, stop and request re-plan via
requirements-to-plan
Validation Requirements
Validation is mandatory at both batch and final stages.
- Run relevant tests for changed scope
- Run lint/build checks where applicable
- Verify acceptance criteria mapping
- Verify no unrelated regressions observed in touched areas
- If validation cannot run, state exactly why and mark risk
Required Outputs
Produce all sections below.
1) Execution Summary
- Scope executed
- Batches completed
- Deviations encountered
2) Batch Logs
For each batch:
- Tasks completed
- Files/modules changed
- Validation performed
- Outcome
3) Acceptance Criteria Trace
Map each acceptance criterion to:
- Implemented change
- Validation evidence
- Status (met/partial/not met)
4) Risk and Issue Register
- New risks discovered
- Open issues
- Temporary compromises (if any)
5) Closeout
- Docs updated (feature/module/system docs as needed)
docs/known-gaps.mdupdated with unresolved unknowns- Operational notes (migration flags, config toggles, rollout notes)
- Final recommendation (ready/not ready)
Hard Gates
- Do not expand scope silently
- Do not skip validation without explicit risk callout
- Do not claim completion with unmet acceptance criteria
- Do not bypass rollback considerations for high-risk changes
Closeout Rules
When execution changes architecture, behavior, or flow semantics:
- Update relevant docs under
/docs - Keep observed vs inferred vs unknown labeling in new analysis notes
- Record contradictions between plan and implementation outcomes
Quality Checklist
- Every completed task maps to a planned item
- Every acceptance criterion has evidence
- Any unmet criterion has a concrete follow-up action
- Final output is decision-ready for release review
More from essentialsoft/agentskills
git-commit-push
create git commits and push code to GitHub safely. use when the user asks to commit, create branch, push changes, or prepare a pull request-ready commit with clear scope and verification.
11codebase-doc-writer
analyze an existing code repository and its current docs to generate and maintain system documentation, component maps, runtime flow indexes, module indexes, feature indexes, ai context, and known gaps. use when the user wants to understand how a program works, document architecture, trace implementation paths, prepare future maintenance context, or refresh project documentation after changes.
4git-commit
Execute git commit with conventional commit message analysis, intelligent staging, and message generation. Use when user asks to commit changes, create a git commit, or mentions "/commit". Supports: (1) Auto-detecting type and scope from changes, (2) Generating conventional commit messages from diff, (3) Interactive commit with optional type/scope/description overrides, (4) Intelligent file staging for logical grouping
2writing-plans
Use when you have a spec or requirements for a multi-step task, before touching code
1brainstorming
You MUST use this before any creative work - creating features, building components, adding functionality, or modifying behavior. Explores user intent, requirements and design before implementation.
1executing-plans
Use when you have a written implementation plan to execute in a separate session with review checkpoints
1