research-deliver

Validate the leading solution concept against evidence and success criteria and turn it into a final recommendation with action items and risks. Use when a preferred solution has emerged and the user

SKILL.md

GitHub

Deliver

Converge on the final recommendation for "$ARGUMENTS". Validate the top solution concept(s) with evidence, refine the approach, and produce actionable recommendations.

Arguments. $ARGUMENTS may contain the project name and optional flags. Take the first token as <project> and treat --skip-checks as a flag; never include flags in file paths. All paths below use <project>. If --skip-checks is present, skip the prerequisite check and, before writing any artifact, append one line to .singularskills/log.jsonl: {"ts":"<ISO timestamp>","skill":"research-deliver","workflow":"research","phase":4,"project":"<project>","event":"skipped","artifact":"04-deliver.md","note":"<reason or \"none given\">"}.

Prerequisites

Part of a workflow. This skill is phase 4 (Deliver) of the Double Diamond workflow (research); cross-domain workflows run it too. First check .singularskills/log.jsonl for an open run for this project: a run-started line with "project":"<project>" and no later run-finished line for the same workflow and project. If one exists, the conductor is running you: continue here. Otherwise, if the user asked to start, run, continue or resume a process or workflow rather than this one step, do not continue here: invoke the workflow skill with the workflow the user asked for, or research when they named none, followed by <project>; it runs this phase in order, asks before it writes and records the run. Otherwise continue here: a later phase runs on its own, behind its gate.

Run this skill's gate before writing: sh <this skill's folder>/scripts/gate.sh <project> <project root>, where <project root> is the folder that contains .singularskills/ (use . when you are already there). If it exits non-zero, stop and tell the user what it printed, unless --skip-checks was given (see Arguments). On a refusal, before you stop, append this line to .singularskills/log.jsonl (create the folder and file if missing): {"ts":"<ISO timestamp>","skill":"research-deliver","workflow":"research","phase":4,"project":"<project>","event":"refused","artifact":"04-deliver.md","note":"<what the gate printed>"}. Once the gate has passed, or under --skip-checks once you have appended the skipped line, append this line before you start work: {"ts":"<ISO timestamp>","skill":"research-deliver","workflow":"research","phase":4,"project":"<project>","event":"started","artifact":"04-deliver.md","note":""}.

Read .singularskills/research/<project>/01-discover.md, .singularskills/research/<project>/02-define.md, and .singularskills/research/<project>/03-design.md.

If 03-design.md does not exist, tell the user:

Phase 3 (Design) has not been completed for "<project>". Run /research-design <project> first, or use --skip-checks to bypass.

Process

  1. Read all prior artifacts:

    • Read .singularskills/research/<project>/01-discover.md for original research
    • Read .singularskills/research/<project>/02-define.md for problem statement and success criteria
    • Read .singularskills/research/<project>/03-design.md for recommended solutions and validation plan
  2. Validate the top solution(s):

    • Use WebSearch to find evidence supporting or contradicting the recommended approach
    • Use WebFetch to read detailed case studies of similar implementations
    • Look for: success stories, failure modes, edge cases, unintended consequences
    • Test each assumption identified in the Design phase
  3. Assess against success criteria:

    • Review each success criterion from the Define phase
    • For each criterion, document evidence that the solution meets it (or doesn't)
    • Flag any criteria that can't be validated with desk research alone
  4. Refine recommendations:

    • Adjust the solution based on validation findings
    • Add implementation considerations discovered during validation
    • Identify quick wins vs. long-term investments
    • Note dependencies and prerequisites
  5. Write the artifact to .singularskills/research/<project>/04-deliver.md with sections:

    • Executive Summary — one-paragraph recommendation
    • Validation Findings — evidence for/against the solution, organized by assumption
    • Success Criteria Assessment — each criterion with evidence status
    • Final Recommendation — refined solution with implementation approach
    • Action Items — prioritized next steps with owners/timelines if applicable
    • Risks & Mitigations — remaining risks and how to handle them
    • Research Journey — brief timeline of what was discovered across all 4 phases

Output

The final deliverable written to .singularskills/research/<project>/04-deliver.md. Present a summary to the user highlighting:

  • The final recommendation (1-2 sentences)
  • Confidence level based on evidence strength
  • Top 3 action items
  • Key risks to watch

Record the run. After you have written the artifact, append this line to .singularskills/log.jsonl (the started line of Prerequisites, with "event":"finished"): {"ts":"<ISO timestamp>","skill":"research-deliver","workflow":"research","phase":4,"project":"<project>","event":"finished","artifact":"04-deliver.md","note":""}.

Open the artifact with this frontmatter (agent is the agent you run in, or unknown; gate is passed when the gate ran clean and skipped under --skip-checks):

---
skill: research-deliver
workflow: research
phase: 4
project: <project>
started: <ISO timestamp>
finished: <ISO timestamp>
agent: <the agent you run in, or "unknown">
gate: <passed|skipped>
---

End the artifact with ## Decisions and ## Open questions, in this shape:

## Decisions
- <what was chosen> — <why> — <what was rejected>

## Open questions
- <what the next phase or a reviewer must settle>