agentsclimarketplace

Technical documentation

Skill shinzoxD/knackbox/skills/documents/technical-documentation

Curated, auditable, benchmark-ready Agent Skills library for Claude Code, Codex, OpenCode, Cursor, and more.

Install
npx -y skills add shinzoxD/knackbox --skill technical-documentation

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

2 things to look at

  • 27 days oldThe repository was created 27 days ago. New is not bad, but a brand new repository carrying a familiar-sounding name is the shape a typosquat arrives in, and there has been no time for anyone else to find a problem with it.
  • 0 stars0 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

What its author says it does

Copied from the file, not written here

Write and improve technical documentation, tutorials, runbooks, and reference pages that users can follow successfully. Use whenever the user asks for docs, a README, setup instructions, an operational runbook, API usage guidance, or an explanation of how a system works, including requests to document code.

The file declares its own license as Apache-2.0. That is the author’s claim about this one file, and it is not the same thing as the license GitHub reports for the repository, which is listed with the other numbers below.

SKILL.md

2.6 KB, as published. Nobody here has run it

Technical Documentation

Write for a named audience completing a concrete task. Documentation is correct only when its commands, prerequisites, expected results, and recovery paths agree with the system.

Workflow

  1. Identify audience knowledge, task, environment, and supported versions.
  2. Choose the document type: tutorial, how-to, explanation, reference, or runbook. Do not mix all types into one page without clear sections.
  3. Verify facts from source code, configuration, or authoritative references.
  4. Put prerequisites before the first action.
  5. Write ordered steps with commands and expected results.
  6. Add verification, rollback, troubleshooting, and ownership where relevant.
  7. Test commands or label anything that was not executed.

Standard structures

How-to

# <Task>
## Prerequisites
## Steps
## Verify
## Troubleshoot
## Undo

Runbook

# <Incident or operation>
**Trigger:** When to use this runbook
**Impact:** What users experience
**Owner:** Team or role
## Diagnose
## Mitigate
## Recover
## Verify
## Escalate

Rules

  1. Use exact file paths, commands, values, and UI labels.
  2. Put one action per numbered step and its expected result immediately after it.
  3. Explain destructive commands before showing them and provide a safer preview when possible.
  4. Define unfamiliar terms once; do not expand standard terms repeatedly.
  5. Keep examples internally consistent and free of real secrets.
  6. Link to source-of-truth reference material instead of duplicating volatile tables or configuration lists.
  7. State the last verified version or date for operational instructions.

Edge cases

  • If repository behavior conflicts with existing docs, flag the conflict rather than choosing one silently.
  • If commands cannot be tested, mark them unverified and identify prerequisites for verification.
  • For multiple platforms, separate platform-specific steps instead of embedding dense conditional instructions in every line.

Keep looking

Skills are one crate of 328,083. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.