Skip to content
Back to skills

Report Writing

ASecurity

Write and deliver the final research report in the fixed factory shape. Use when your research is done, when you are about to run `factory report` or `factory done`, or when asked for findings.

  • 8 stars
  • 0 votes
  • 0 copies
  • 0 views
  • Added November 5, 2026
ai-agentsgo

Security analysis

A100/100

Scanned November 5, 2026

npx -y skills add volekerb/hivemill --skill report-writing --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Report Writing?

Add the live security badge to your README. It updates with every re-scan.

Security grade badge for Report Writing
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/volekerb-report-writing/badge)](https://www.skillsdirectory.com/skills/volekerb-report-writing)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

Download with Pro
SKILL.md
---
name: report-writing
description: Write and deliver the final research report in the fixed factory shape. Use when your research is done, when you are about to run `factory report` or `factory done`, or when asked for findings.
---

Purpose: a report an owner can read in one minute, with every claim traceable to a source.

Follow these steps in order.

1. Use exactly these headings, in this order:
   - `## Question`
   - `## Answer`
   - `## Findings`
   - `## Recommendation`
   - `## Proposed follow-ups`
   - `## Evidence`
   - `## Limits`
2. Question: the question as you understood it, one or two lines, plus the brief note and the hypotheses (H1, H2, ...).
3. Answer: a plain-English summary for the owner, at most 10 lines, no jargon, no source ids. It answers the question first.
4. Findings: numbered. Each finding is one claim, says which hypothesis it supports or refutes, notes the counter-evidence sought and found, and ends with its source ids, like `[S1]` or `[S2][S4]`. No finding without a source.
5. Recommendation: what to do, in order, with adopt, adapt or skip where it applies.
6. Proposed follow-ups: tasks you suggest, one per line. You do not create tasks.
7. Evidence: numbered `[Sn]` entries. Each has the URL or command, the date accessed, and primary or secondary. Include raw command output for measurements.
8. Limits: what you did not check, single-source claims, the time box, anything uncertain, and one `Verified:` line: what you re-checked and what changed.
9. Verify before you publish: re-open the cited source of each finding and re-check every number, version, date and quote against it; re-run cheap measurement commands; check the Answer against the Findings. Fix or drop what does not match; list what you could not re-check under Limits.
10. Store the full report: `factory task note <own task> -` with the report as the heredoc body.
11. Publish it: `factory report "<one-line answer>" - --goal <G>` with the same body.
12. Finish: `factory done "<one-line answer>; report in task note"`.

Stop and ask your lead when: the answer does not follow from the evidence, or the question changed since the brief.

Attribution

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

View all in ai-agents →

Comments

Loading comments…

Newsletter

Get the weekly list of safe new skills, and the ones we pulled.