Skip to content
Back to skills

Investigation

ASecurity

Answer a read-only engineering question from repository evidence without changing implementation files.

  • 2 stars
  • 0 votes
  • 0 copies
  • 0 views
  • Added October 1, 2026
ai-agents

Security analysis

A100/100

Scanned October 6, 2026

npx -y skills add williamwue/oh-my-stack --skill investigation --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Investigation?

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

Security grade badge for Investigation
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/williamwue-investigation/badge)](https://www.skillsdirectory.com/skills/williamwue-investigation)

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: investigation
description: "Answer a read-only engineering question from repository evidence without changing implementation files."
disable-model-invocation: true
---

# Investigation

## Child session handoff

Read [the handoff contract](../poteto-mode/references/subagent-handoff.md).
New tasks, repair rounds, retries, and queue items use fresh child sessions
with the original brief, every later directive, prior findings and responses,
and unresolved objections. Reuse only for required costly live state, and
only when the host allows it. Stop and fence active writers before replacement.
A host-owned orchestrator's model catalog, workspace binding, child tools,
and review-round rules take precedence over the native binding above.
Keep its task handles and attribution receipts. Do not use a backing child
conversation as a new delegated review, or claim native-record verification
for a host-owned child. Report attribution evidence gaps explicitly.

Use this workflow for code behavior, ownership, layering, feasibility, and
engineering tradeoff questions whose requested outcome is an answer rather
than a change.

1. Freeze the question, inspected scope, and current revision when available.
2. Run the `how` workflow over the affected area. For an alternatives decision,
   gather evidence for each viable option under the same constraints.
3. Verify important claims against current files, executable behavior, or both.
4. State facts, inferences, and unknowns separately. Do not infer historical
   motivation without evidence.
5. Keep the workspace unchanged. If investigation reveals a requested change,
   stop at the diagnosis and route a later implementation request separately.
6. Apply `technical-writing` and `unslop` when producing a durable document.

Return the `how` output for explanatory questions. For a decision, return the
decision criteria, evidence, tradeoffs, recommendation, and verification
boundary. State that the investigation was read-only.

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.

Comments

Loading comments…