Report

Generate a cross-domain report for "$ARGUMENTS" (or every project under .singularskills/ if no argument is given), built from the run log, the artifacts' frontmatter and their Decisions and Open questions sections.

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":"singular-report","workflow":"singular","phase":null,"project":"<project>","event":"skipped","artifact":"<artifact file>","note":"<reason or \"none given\">"}.

Prerequisites

Requires a .singularskills/ folder with at least one artifact (for <project>, when given). If there is none, stop and tell the user that nothing has been saved yet.

Process

1. Read the workflows and the format. Read every *.yml in the workflows/ folder next to the skills/ folder of this pack; if it is not there, read the copies of the workflow files bundled with the workflow skill (the conductor). The log and artifact format is described in the run-log.md reference bundled with the workflow skill. Read .singularskills/config.md if it exists for project context.

2. Read the run log. If .singularskills/log.jsonl exists, keep the lines for <project> (or for every project): phase-approved lines (which phases a person approved, and when), skipped and refused lines (gates bypassed or refused, with their notes), and run-started / run-finished lines (runs still open).

3. Read the artifacts. For each .singularskills/<folder>/<project>/ folder (skip reports/), read every artifact:

   • its frontmatter: skill, workflow, phase, started, finished, agent, gate;
   • its Decisions and Open questions sections — the closing ## Decisions / ## Open questions of a single-writer artifact, or each member's ### Decisions / ### Open questions inside its ## <Skill title> section of a shared artifact;
   • its ## Recalled section (or a member's ### Recalled), when a first-phase artifact has one: what the skill found about the project before it asked anything;
   • its ## Tools section, when present: which Figma capabilities were bound to which tools, or that no server was connected; a utility artifact (frontmatter phase: null, no NN- prefix) is listed after the phases;
   • only as much of the body as you need to explain a decision.

   Give each phase a state: approved (logged approval), written (artifact present, no approval) or missing, as the run-log.md reference bundled with the workflow skill defines.

4. Summarise per workflow and domain: phases approved, written and missing; the key decisions (what was chosen, why, what was rejected); the open questions still standing; every skipped gate (frontmatter gate: skipped or a skipped log line), stated plainly.

5. Identify cross-domain connections, using the decisions themselves:

   • Do the design decisions follow the research findings?
   • Is the branding followed in marketing?
   • Are accessibility issues addressed in development?
   • Do open questions in one domain block a phase in another?

6. Assess readiness: what is approved and can be relied on, what is written but awaits review, what is missing or blocked, and which domains have not been started but should be.

7. Write the report to .singularskills/reports/<project>-report.md (or .singularskills/reports/<date>-project-report.md if no argument), with sections:

   • Executive Summary — one paragraph on where the project stands
   • Workflow Status — per workflow: each phase's state, approval date and gate
   • Key Decisions — per domain, quoted or condensed from the Decisions sections, each attributed to its artifact
   • Open Questions — collected from the artifacts, grouped by who or which phase must settle them
   • Cross-Domain Insights — connections and conflicts
   • Readiness Assessment — approved, awaiting review, missing or blocked
   • Recommended Next Steps — prioritized actions, naming the workflow and phase to run next

Output

The report written to .singularskills/reports/. Present the executive summary and recommended next steps to the user.