MCAF: Feature Spec
Trigger On
- add or change non-trivial behaviour
- behaviour is under-specified and engineers are guessing
- tests need a stable behavioural source of truth
Value
- produce a concrete project delta: code, docs, config, tests, CI, or review artifact
- reduce ambiguity through explicit planning, verification, and final validation skills
- leave reusable project context so future tasks are faster and safer
Do Not Use For
- architecture decisions that need alternatives and trade-offs
- tiny typo or cosmetic-only changes with no behavioural impact
Inputs
docs/Architecture/Overview.md
- the nearest
AGENTS.md
- current user flows, business rules, and acceptance expectations
Quick Start
- Read the nearest
AGENTS.md and confirm scope and constraints.
- Run this skill's
Workflow through the Ralph Loop until outcomes are acceptable.
- Return the
Required Result Format with concrete artifacts and verification evidence.
Workflow
- Define scope first: in scope, out of scope, boundaries touched.
- If the feature doc is missing, scaffold from
references/feature-template.md.
- Keep the spec executable:
- numbered rules
- main flow
- edge and failure flows
- system behaviour
- verification steps
- Definition of Done
- Make the spec concrete enough that tests can be written without guessing.
- If the feature creates a new dependency, boundary, or major policy shift, update an ADR too.
Deliver
docs/Features/feature-name.md
- a feature spec that engineers and agents can implement directly
Validate
- rules are testable, not aspirational
- edge cases are captured where they matter
- verification steps match the intended behaviour
- the doc can drive implementation without hidden tribal knowledge
Ralph Loop
Use the Ralph Loop for every task, including docs, architecture, testing, and tooling work.
- Plan first (mandatory):
- analyze current state
- define target outcome, constraints, and risks
- write a detailed execution plan
- list final validation skills to run at the end, with order and reason
- Execute one planned step and produce a concrete delta.
- Review the result and capture findings with actionable next fixes.
- Apply fixes in small batches and rerun the relevant checks or review steps.
- Update the plan after each iteration.
- Repeat until outcomes are acceptable or only explicit exceptions remain.
- If a dependency is missing, bootstrap it or return
status: not_applicable with explicit reason and fallback path.
Required Result Format
status: complete | clean | improved | configured | not_applicable | blocked
plan: concise plan and current iteration step
actions_taken: concrete changes made
validation_skills: final skills run, or skipped with reasons
verification: commands, checks, or review evidence summary
remaining: top unresolved items or none
For setup-only requests with no execution, return status: configured and exact next commands.
Load References
- use
references/feature-template.md only for scaffolding
Example Requests
- "Write a feature spec for the new checkout retry flow."
- "Document the behaviour before coding this API change."
- "Turn this loose requirement into an executable feature doc."
1---2name: mcaf-feature-spec-43description: "Create or update a feature spec under `docs/Features/` with business rules, user flows, system behaviour, verification, and Definition of Done. Use when the user asks for a feature spec, executable requirements, acceptance criteria, behaviour documentation, or a pre-implementation plan for non-trivial behaviour changes."4compatibility: "Requires repository write access; produces Markdown docs with Mermaid diagrams and executable verification steps."5metadata:6 source: https://skillmd.com/skills/managedcode/mcaf-feature-spec-47---89# MCAF: Feature Spec1011## Trigger On1213- add or change non-trivial behaviour14- behaviour is under-specified and engineers are guessing15- tests need a stable behavioural source of truth1617## Value1819- produce a concrete project delta: code, docs, config, tests, CI, or review artifact20- reduce ambiguity through explicit planning, verification, and final validation skills21- leave reusable project context so future tasks are faster and safer2223## Do Not Use For2425- architecture decisions that need alternatives and trade-offs26- tiny typo or cosmetic-only changes with no behavioural impact2728## Inputs2930- `docs/Architecture/Overview.md`31- the nearest `AGENTS.md`32- current user flows, business rules, and acceptance expectations3334## Quick Start35361. Read the nearest `AGENTS.md` and confirm scope and constraints.372. Run this skill's `Workflow` through the `Ralph Loop` until outcomes are acceptable.383. Return the `Required Result Format` with concrete artifacts and verification evidence.3940## Workflow41421. Define scope first: in scope, out of scope, boundaries touched.432. If the feature doc is missing, scaffold from `references/feature-template.md`.443. Keep the spec executable:45 - numbered rules46 - main flow47 - edge and failure flows48 - system behaviour49 - verification steps50 - Definition of Done514. Make the spec concrete enough that tests can be written without guessing.525. If the feature creates a new dependency, boundary, or major policy shift, update an ADR too.5354## Deliver5556- `docs/Features/feature-name.md`57- a feature spec that engineers and agents can implement directly5859## Validate6061- rules are testable, not aspirational62- edge cases are captured where they matter63- verification steps match the intended behaviour64- the doc can drive implementation without hidden tribal knowledge6566## Ralph Loop6768Use the Ralph Loop for every task, including docs, architecture, testing, and tooling work.69701. Plan first (mandatory):71 - analyze current state72 - define target outcome, constraints, and risks73 - write a detailed execution plan74 - list final validation skills to run at the end, with order and reason752. Execute one planned step and produce a concrete delta.763. Review the result and capture findings with actionable next fixes.774. Apply fixes in small batches and rerun the relevant checks or review steps.785. Update the plan after each iteration.796. Repeat until outcomes are acceptable or only explicit exceptions remain.807. If a dependency is missing, bootstrap it or return `status: not_applicable` with explicit reason and fallback path.8182### Required Result Format8384- `status`: `complete` | `clean` | `improved` | `configured` | `not_applicable` | `blocked`85- `plan`: concise plan and current iteration step86- `actions_taken`: concrete changes made87- `validation_skills`: final skills run, or skipped with reasons88- `verification`: commands, checks, or review evidence summary89- `remaining`: top unresolved items or `none`9091For setup-only requests with no execution, return `status: configured` and exact next commands.9293## Load References9495- use `references/feature-template.md` only for scaffolding9697## Example Requests9899- "Write a feature spec for the new checkout retry flow."100- "Document the behaviour before coding this API change."101- "Turn this loose requirement into an executable feature doc."