Exécutez n'importe quel Skill dans Manus
en un clic

Exécutez n'importe quel Skill dans Manus en un clic

$pwd:

investigate

Name: Investigate
Author: blueberrycongee

// Systematic debugging skill. Use when encountering a bug, test failure, unexpected behavior, or when asked to "investigate", "debug", "diagnose", or "figure out why". Enforces root-cause-first discipline with structured hypothesis tracking.

Exécuter dans Manus

$ git log --oneline --stat

stars:303

forks:21

updated:31 mars 2026 à 07:35

SKILL.md

readonly

name	investigate
description	Systematic debugging skill. Use when encountering a bug, test failure, unexpected behavior, or when asked to "investigate", "debug", "diagnose", or "figure out why". Enforces root-cause-first discipline with structured hypothesis tracking.

Investigate

Systematic debugging workflow. Do not guess-and-fix — find the root cause first.

Phase 1: Collect

Gather all available evidence before forming any hypothesis.

Read the error message, stack trace, or symptom description exactly as given
Identify the affected code path — read the relevant source files
Check git log --oneline -20 in the affected area for recent changes
If reproducible, reproduce the issue and capture the exact output
List what you know and what you do not know

Deliverable: a structured symptom summary with affected files, error output, and recent changes.

Phase 2: Hypothesize

Form up to 3 ranked hypotheses. For each:

State the hypothesis in one sentence
Identify the single cheapest check that would confirm or refute it
Do NOT start fixing yet

Phase 3: Verify

Test hypotheses in rank order. For each:

Run the cheapest check identified in Phase 2
If confirmed — move to Phase 4
If refuted — record what you learned and move to the next hypothesis

3-strike rule: If all 3 hypotheses fail, step back and re-collect. Do not keep guessing. Re-read the code path more carefully, expand the search area, or check for environmental factors (config, dependencies, platform).

Phase 4: Fix

Only after root cause is confirmed:

Write the minimal fix that addresses the root cause
Do not refactor surrounding code
Do not add speculative defensive checks
Run the test suite to verify the fix and check for regressions

Phase 5: Verify and Report

Confirm the original symptom is resolved
Confirm no regressions were introduced
Summarize: root cause, fix applied, evidence of resolution

Rules

Never skip Phase 1-3 and jump to fixing
Never change tests to match broken behavior
If the bug is in a dependency or environment, report it — do not patch around it silently
If you cannot reproduce the issue, say so and explain what you tried

related-skills.json

même dépôt

code-review.md

from "blueberrycongee/termcanvas"

Structured code review skill. Use when asked to "review this code", "review this PR", "check this diff", or when acting as a Hydra reviewer. Runs a multi-pass review with specialist focus areas and confidence-gated findings.

2026-04-12303

qa.md

from "blueberrycongee/termcanvas"

QA testing skill with real browser automation. Use when asked to "test this site", "QA this page", "check for visual bugs", "verify the deploy", or when Hydra needs browser validation for UI changes. Requires the browse binary.

2026-04-12303

using-termcanvas.md

from "blueberrycongee/termcanvas"

Use when starting work in a TermCanvas-managed repo to route between direct work, Hydra, or a narrow TermCanvas skill.

2026-04-12303

hydra.md

from "blueberrycongee/termcanvas"

Use when a task should run through Hydra's Lead-driven workflow for multi-agent orchestration, or when an existing workflow must be inspected or cleaned up.

2026-04-11303

challenge.md

from "blueberrycongee/termcanvas"

Adversarial review skill. Use when the user wants to stress-test an idea, argument, proposal, or opinion from multiple independent angles. Spawns parallel Hydra workers with orthogonal analytical methodologies.

2026-04-08303

security-audit.md

from "blueberrycongee/termcanvas"

Security audit skill. Use when asked to "audit security", "check for vulnerabilities", "security review", "pentest", or when evaluating code that handles auth, user input, secrets, or external data. Runs a phased scan covering OWASP Top 10 and STRIDE threat modeling.

2026-03-31303

package.json