Skip to content

dongshuyan/compass-skills / skills/run-history-skill-builder

run-history-skill-builder

Verified

Turn a completed task, browser flow, artifact pipeline, failure-recovery trace, or repeatedly refined workflow into a new reusable skill package or a reviewed skill-design plan. Use when the user asks to make a new skill from real run history, extract a reusable workflow from conversation/logs/files, summarize lessons into a new skill, or produce a plan before writing files. Do not use to upgrade an existing skill or to execute the business workflow itself.

references/examples.md

View on GitHub
# Examples

These examples are public, sanitized patterns. They exist to teach reusable packaging decisions, not to preserve one private machine or one one-off task.

## Example 1: Browser Publishing Workflow

User request:

```text
Turn the browser-based publishing workflow we just stabilized into a new skill.
```

Evidence worth extracting:

- a validated final artifact exists and is distinct from raw intermediates;
- the form has required metadata fields and a risky final submit step;
- there was a false-success failure mode where a visible field looked correct but the submitted value was wrong;
- success must be proven by a platform-visible confirmation, resulting artifact, or other checkable evidence.

Good package outcome:

- state gates for input resolution, metadata readiness, pre-submit validation, submission, and success verification;
- a failure-recovery example for the false-success case;
- privacy boundaries that keep account data and credentials out of the package.

## Example 2: Local Document-Conversion Pipeline

User request:

```text
We finally stabilized a DOCX to PDF workflow. Make it a reusable skill.
```

Package outcome:

- input and output naming contract;
- deterministic conversion or validation helpers in `scripts/`;
- checks for output existence, page count, extraction sanity, and common failure states;
- examples for missing fonts, layout drift, or failed generation.

## Example 3: Plan-Only Design

User request:

```text
Give me a full skill-design plan, but do not write files yet.
```

Expected behavior:

- list the candidate files and their owners;
- explain the validation plan;
- explicitly say that no files were written.

## Example 4: Existing Skill Upgrade Handoff

User request:

```text
Improve the existing release-publisher skill based on this failure.
```

Expected behavior:

- recognize that this is not a new-skill request;
- prepare a clean handoff for `$run-history-skill-upgrader`;
- do not create a new skill package and do not edit the existing skill here.

## Example 5: Minimal Package

User request:

```text
Package this short local query workflow as a skill.
```

Expected behavior:

- first decide whether the workflow is frequent enough, error-prone enough, and stable enough to deserve a skill;
- if yes, prefer only `SKILL.md` plus optional `agents/openai.yaml`;
- avoid empty `references/`, `scripts/`, or `evals/`.

## Example 6: Pattern-Mining Request

User request:

```text
Search similar public skills, inspect them, and then package our workflow.
```

Expected behavior:

- inspect real public references and record what was adopted;
- do not cite repositories you did not read;
- do not ship the temporary research clones as part of the released skill.
writing-skillsobra/superpowersUse when creating new skills, editing existing skills, or verifying skills work before deploymentStars: 322kUpdated Claude CodeCodexVerifiedInstall
xlsxanthropics/skillsUse this skill any time a spreadsheet file is the primary input or output. This means any task where the user wants to: open, read, edit, or fix an existing .xlsx, .xlsm, .xltx, .csv, or .tsv file (e.g., adding columns, computing formulas, formatting, charting, cleaning messy data); create a new spreadsheet from scratch or from other data sources; or convert between tabular file formats. Trigger especially when the user references a spreadsheet file by name or path — even casually (like "the xlsx in my downloads") — and wants something done to it or produced from it. Also trigger for cleaning or restructuring messy tabular data files (malformed rows, misplaced headers, junk data) into proper spreadsheets. The deliverable must be a spreadsheet file. Do NOT trigger when the primary deliverable is a Word document, HTML report, standalone Python script, database pipeline, or Google Sheets API integration, even if tabular data is involved.Stars: 198kUpdated Claude CodeCodexVerifiedInstall
shipping-and-launchaddyosmani/agent-skillsPrepares production launches. Use when preparing to deploy to production, or when asking what needs to be in place before shipping. Use when you need a pre-launch checklist, when setting up monitoring, when planning a staged rollout, or when you need a rollback strategy.Stars: 119kUpdated Claude CodeCodexVerifiedInstall
childhoodthedotmack/claude-memBabysit a shipped release after merge, publication, or deployment. Verify published packages, production behavior, installed plugins, and late CI or user reports until the observation window is clean; use babysit for PR readiness before release.Stars: 118kUpdated Claude CodeCodexVerifiedInstall