This repository packages the Unknowns Workflow as portable agent skills and templates for agent-assisted implementation.
- Keep the project generic. Do not add private workspace names, personal policies, account details, secrets, or organization-internal process names.
- Treat
skills/as the installable skill source of truth. - Treat
docs/andtemplates/as public-facing explanation and reusable artifacts. - Preserve the core workflow: classify unknowns before implementation, capture plan deviations during implementation, and turn the results into learning artifacts after implementation.
- Do not publish credentials, raw private logs, screenshots containing private information, or local runtime output.
The portable layer is the eight-skill, six-template workflow and its docs; it
works without Python. The optional unknowns_harness layer is a local,
non-authoritative observer and does not grant authority or perform actions.
Do not change without a recorded decision:
- package, module, function, and dataclass field names;
- console script names, flags, exit codes, JSON keys, and key ordering;
- typed error strings and rejection fields;
- the eight skills and six templates, including their filenames and content.
The exact module map and platform boundary are in docs/architecture.md.
evidence_graph, consultation_budget, and semantic_fingerprint use
type(value) is not str deliberately; isinstance(True, int) is True.
The harness observes local inputs only. It must not grant authority, issue tokens or permissions, perform external actions, or add an authority layer.
Before pushing, run:
python3 -m unittest discover -s tests -q
python3 -m json.tool .agents/plugins/marketplace.json >/dev/null
python3 -m json.tool .codex-plugin/plugin.json >/dev/null
git diff --checkWhen changing Markdown, inspect the rendered structure or at least review the
headings and links. Read docs/architecture.md for the complete map and test
coverage summary.