agentleFS
pchemguypchemguy/Skill-SDD-Manager/skills/sdd-verify/SKILL.md

sdd-verify

Skill0 starsChanged 37 days agoMITView on GitHub

Use when selecting and running checks for an SDD task, selected change, milestone, phase, or project; assessing acceptance and exit-condition coverage; classifying observed failures; performing read-only milestone/phase implementation code review, or returning verification evidence and remaining gaps before completion.

5 other projects use a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

1 rule. 0 are common, 1 is distinctive, and 0 are very specific to this project.

  • 0

    Common

    Most of the 5 edited versions we read keep these.

  • 1

    Distinctive

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Run the selected checks and inspect actual outcomes, including failures, warnings, skips, and collection counts where relevant.

    line 23

“Common” is measured: kept by at least 6 in 10 of the 5 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Verify a selected work boundary

Establish the requested task, change, milestone, phase, or project scope. Load the relevant references:

WorkLoad
Review milestone/phase implementation code and findingsboundary review
Derive checks from requirements, changes, and dependenciescheck selection
Execute checks and assess acceptance evidenceexecution and evidence
Classify failures and return unresolved workfailure assessment

Use a current sdd-orient handoff for governing instructions, Git state, dirty-path ownership, relevant documents, and declared commands. sdd-manage coordinates the requested scope and execution prerequisites; the active implementation workflow, sdd-implement for main task work or sdd-steer for a checkpoint amendment, requests verification within its boundary. Read the owning TASKS or FEATURE-TASKS, relevant SPEC contracts, PLAN exit conditions, design and layout, actual changes, tests, and existing evidence as needed. Read-only check planning can precede execution; commands with side effects require the established eligible Git worktree and authorized environment.

For an explicit boundary review task, load boundary review and assess implementation code separately from test execution. Neither activity substitutes for the other. Return findings and recheck evidence without repairs or task/hosted state changes.

  1. Identify the acceptance and exit conditions to verify and the implementation state being checked.
  2. Select sufficient direct, dependent, integration, and boundary checks. Use project commands and an existing verification map where useful; neither a map nor new scripts are required.
  3. Run the selected checks and inspect actual outcomes, including failures, warnings, skips, and collection counts where relevant.
  4. Associate each condition with evidence and any limitation. Distinguish verified, failed, blocked, and not checked conditions; passing checks do not establish acceptance they never exercised.
  5. Classify observed failures where evidence permits and return remaining work to its owner. Do not silently omit an unrelated or pre-existing failure from the results.

sdd-tdd owns testing strategy and test design or changes; this skill assesses coverage and executes the verification campaign. sdd-docs owns documentation maintenance. The active implementation workflow owns repairs, completion checkboxes, commits, pushes, and issue-closure coordination. Leave source, tests, governing documents, and task status unchanged during verification; report necessary changes rather than repairing them here. Account for command-generated artifacts and preserve unrelated pending work.

Return the boundary and implementation state, conditions assessed, check selection and rationale, commands and environment, observed outcomes and evidence locations, failure classifications, omitted or blocked checks, and remaining gaps. sdd-report owns presentation of these facts. Existing task evidence or a project-designated record may hold durable results; no separate verification journal is required. State when the selected conditions are verified without equating that finding with task completion or whole-project correctness.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.