Collect task context
Skill Mergemat/agent-readiness-skills/skills/collect-task-context
Harness-agnostic skills for agent-readable repos and task prep
npx -y skills add Mergemat/agent-readiness-skills --skill collect-task-contextAssembled from the repository path, not quoted from the project. Check it against their README if it does not work.
2 things to look at
- no licenseNo license file was found in the repository. Code published without one is not open source by default, so using it at work is a question for whoever answers licensing questions where you are.
- 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
Builds a minimal task-specific context packet for an agent. Use when user wants context for a coding task, needs to prepare an agent handoff, or asks what files, constraints, tests, and evidence should be given to an implementation agent.
SKILL.md
1.6 KB, 279 tokens by cl100k_base, as published. Nobody here has run it
Collect Task Context
Gather the smallest reproducible context packet for one task.
This is not a repo dump. Include only evidence that helps plan, edit, or verify the task.
Process
- Identify the task, expected behavior, and acceptance criteria.
- Read project agent docs and relevant domain docs.
- Map allowed changes, read-only context, and forbidden paths.
- Find change candidates, call sites, interfaces, tests, fixtures, and configs.
- Record verification commands and known gaps.
- Capture provenance: branch, base, date, and commands used.
Use packet-format.md when writing a formal packet.
Include
- task summary
- relevant files and symbols with why each matters
- scope and forbidden changes
- product/domain rules
- architecture, API, DB, or event contracts when relevant
- focused tests and broader checks
- assumptions and open questions
Exclude
- generated files unless directly relevant
- vendor and caches
- broad docs with no task-specific takeaway
- secrets, tokens, private credentials
- stale docs unless conflict matters
Output
Return ready, ready_with_assumptions, or not_ready.
If not ready, name the blocking missing fact. If ready, keep the packet compact enough that another agent can use it immediately.
What ships with it: 1 file
857 B alongside SKILL.md
references/
- packet-format.md857 B