Skip to content

docs: inline spec 009 fields, drop private cross-reference #3

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
zenprocess merged 1 commit into from
Apr 8, 2026
Merged

docs: inline spec 009 fields, drop private cross-reference #3

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
1 change: 0 additions & 1 deletion .github/workflows/ci.yml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -19,7 +19,6 @@ jobs:
--exclude-loopback
--max-redirects 5
--accept '200..=204,206,301..=308,403,429'
--exclude 'github.com/zenprocess/switchyard'
'./**/*.md'
fail: true

Expand Down
14 changes: 8 additions & 6 deletions README.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -53,14 +53,16 @@ CACP is part of the [standra.ai](https://standra.ai) open standards stack:
- **[Pawbench](https://github.com/zenprocess/pawbench)** — reference benchmark that scores LLMs against CACP-formatted prompts and responses, including the orchestration × complexity matrix.
- **[ServingCard](https://servingcard.dev)** — model serving config standard.

## Recent additive fields (spec 009)
## Additive fields for orchestration & complexity reporting

CACP responses can carry these spec 009 fields (see [switchyard spec 009](https://github.com/zenprocess/switchyard/blob/main/specs/009-pawbench-orchestration-axis/spec.md) for the operational mapping):
CACP responses can carry these additional fields for runners that report multi-dimensional dispatch results. Vocabularies are normative in [Axiom §17](https://github.com/zenprocess/axiom/blob/main/spec.md):

- `complexity_tier` — `display` / `crud` / `transactional` / `cross_cutting`
- `verification_runs[]` — N-run AC re-verification with per-run verdict + prompt hash
- `artifact_quality` — static-analysis score over changed files
- `fixture_gap` (status) — AC un-evaluable due to missing setup, NOT counted against the agent
- `complexity_tier` — `display` / `crud` / `transactional` / `cross_cutting`. Stratifies aggregate scores so display-tier passes don't mask transactional-tier cliffs.
- `verification_runs[]` — N-run AC re-verification, one record per run with `verdict`, `prompt_hash`, `elapsed_ms`. Surfaces verifier flake.
- `artifact_quality` — static-analysis score over the *changed files only*; orthogonal to AC pass.
- `fixture_gap` (terminal status) — AC un-evaluable due to missing setup (seed data, env, services). **Not counted against the agent** — counts against the scenario author.

[Pawbench](https://github.com/zenprocess/pawbench) is the reference implementation of these fields end-to-end.

## License

Expand Down
Loading