To Prompt

What the source does

• FilesystemNone observedNo reads or writes in the source
• NetworkNone observedNo outbound calls in the source
• ShellNone observedNo commands run in the source
• DependenciesPinnedNothing floats
• SecretsNone foundNo keys in source

Five audit layers

Gates applied: no behavioural pass.

Audit 0.4.1 · source sha 118e61cf1c61 · 16 Oct 2026Report an issue or request a re-scan

• Publisherpedronauck GitHub userOwner of the repository this skill is read from
• Source folderpedronauck/skills/skills/mine/to-promptThe exact folder we audited
• LicenceNot declaredAs the repository states it
• Commit pinned for this verdict118e61cf1c61A new audit runs when the source changes
• Repository statusNot archivedLast push 31 days ago
• Sandboxed runNot runSkills are read, not executed; layer L4 is skipped

Its description: Create a handoff brief for another LLM to implement. Preserve accepted decisions and write to docs/prompts/; excludes one-shot questions and end-user copy.

---
name: to-prompt
description: "Create a handoff brief for another LLM to implement. Preserve accepted decisions and write to docs/prompts/; excludes one-shot questions and end-user copy."
disable-model-invocation: true
metadata:
  author: Pedro Nauck
  github: https://github.com/pedronauck
  repository: https://github.com/pedronauck/skills
---
# To Prompt

Write a **brief**, not a solution. The brief carries everything the receiving LLM needs to choose the implementation itself — the problem, the current state, the requirements, the constraints — and preserves any implementation constraints or approved plan already supplied by the user. Leave undecided design choices to the receiver; do not erase accepted decisions to manufacture an open-ended problem.

Every brief is a file: `docs/prompts/<YYYYMMDD-HHmm>_<slug>.md` (e.g. `docs/prompts/20260709-1745_fix-auth-redirect.md`).

## Steps

1. **Classify the task** — bug fix, improvement, or feature — and draft a one-paragraph statement of the problem and the outcome wanted.
2. **Gather the evidence.** Read the code paths involved; quote current code as it exists; capture the relevant error/log excerpts, excluding credentials and unrelated private data; note environment and recent changes. *Done when:* every claim in the brief is backed by something you actually read or ran.
3. **Write the brief from the template.** Copy `assets/brief-template.md` to `docs/prompts/<YYYYMMDD-HHmm>_<slug>.md` (create the directory if absent; slug = lowercase-hyphenated summary; timestamp = now). Fill the applicable sections and remove guidance comments. Record consequential unknowns; omit irrelevant sections.
4. **Check decision fidelity.** Separate user-approved constraints from suggestions and unresolved choices. Preserve the requested outcome and authorization; avoid introducing new implementation mandates.
5. **Hand off.** Reply with the file path and a one-line summary of what the brief asks for.

## Bundled files

- `assets/brief-template.md` — the brief's skeleton and per-type coverage guidance; the single source of truth for what a complete brief contains.

Excerpt, first 2,128 characters of SKILL.md, verbatim from the audited commit. It is text from someone else’s repository, quoted so you can judge it before your agent reads it.

git clone https://github.com/pedronauck/skills
git -C skills checkout 118e61cf1c61

Works in

1. 2026-10-16
   ✓SAFE B 89 / 100

   first audit · source 118e61cf1c61

Questions about To Prompt

• github.com/pedronauck/skills/tree/main/skills/mine/to-prompt — the audited source at commit 118e61cf1c61 and every finding aboveObserved · 16 Oct 2026

Spotted something wrong? Report a correction · Methodology