feat(tooling): changed-scope, telemetry, arch fitness, flake triager, prompt-eval - #21
Merged
Merged
Conversation
…ager, prompt-eval Five shift-left tooling features plus design specs (designs/): - changed-scope (FOUNDRY_SINCE) in ts/java templates + agent-loop; novar folded into java lint via FOUNDRY_BASE_REF; separate no-var CI job removed - loop telemetry: foundry-verb-wrap + foundry-loop-report, wired into both templates and foundry-init; /loop-report skill lives in cmaintz-skills - architecture fitness: presets/arch (architecture-agnostic layer model; dependency- cruiser + ArchUnit templates + coaching guides) - deterministic flake triager: foundry-flaky (gate + classify) + baseline schema - prompt/output eval: presets/prompt-eval (vitest + JUnit ranking harness) - CONTRACT auxiliary-tasks addendum; ruleset_guard gains `lines` + `coverage` kinds Scripts + guard classifiers validated in isolation; live mise/consumer runs pending. Supersedes the stale feat/verb-tiers-wip (a partial first-batch snapshot).
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Tooling suite — five shift-left features (additive)
Additive to foundry: new presets/scripts/mise-tasks + small edits to existing templates. No restructure of foundry internals.
Features
FOUNDRY_SINCE) — ts/java templates run file-local verbs over the changeset at the fast placements; whole-tree in CI (the authoritative backstop). Whole-program checks (typecheck/arch) always run whole-tree. novar folded into javalint, ratchet-scoped viaFOUNDRY_BASE_REF; the separateno-varCI job removed.foundry-verb-wrapemits one JSONL line per verb (timing/exit/scope/placement);foundry-loop-reportsummarises convergence/thrash/time-sinks. Wired into both templates + foundry-init./loop-reportskill ships in the paired cmaintz-skills PR.presets/arch/: an architecture-agnostic layer model (LAYERS + ALLOW + no-cycles), not hexagonal-only — ships hexagonal/layered/clean/modular example maps. dependency-cruiser (TS, underlint) + ArchUnit (Java, undertest) + coaching guides.foundry-flaky(gate + classify over JUnit XML / vitest JSON); quarantine is a ratcheted, labelled decision (never hides a real failure).presets/prompt-eval/: vitest + JUnit ranking harness generalising AutoApplicant's proven pattern (good ≥ floor, weak ≤ ceiling, gap ≥ separation). Scorer stays in the consumer.Also
ruleset_guard.pygainslines(ArchUnit store) +coverage(prompt-eval fixtures — inverse: removal is the loosening) kinds.designs/.Validation
Every script + guard classifier tested in isolation. Live mise/consumer runs (actual ArchUnit/dep-cruiser/flaky/vitest execution + mise-level wiring) are pending a real project — noted per spec. Adoption is opt-in per repo.
Pairs with CMaintz/cmaintz-skills#3 (CONTRACT addendum + /loop-report).