Agent task: bootstrap image lab maturity spine and vision service manifest

[mdheller]

Agent-directed work package.

Target agent: Codex environment for SociOS-Linux/imagelab; Copilot coding agent if available.

Context:
imagelab is the Linux-native lab workspace for image understanding, visual extraction, captioning, chart/diagram analysis, visual classification, and image-to-evidence pipelines for SocioProphet services. It should emit governed functional service manifests and evaluation evidence for downstream platform services.

Scope:

1. Inspect the live repository before editing.
2. Add README explaining purpose, lab-only boundary, and integration points.
3. Add repo.maturity.yaml using repo-maturity.v1 semantics.
4. Add lab.manifest.json covering visual classification, captioning, object/region detection, chart/diagram extraction, image evidence extraction, and multimodal handoff to Holmes.
5. Add service-manifest/functional-service.v1.json conforming to SocioProphet/functional-model-surfaces semantics.
6. Add datasets/README.md, evals/README.md, adapters/README.md, training-runs/README.md, and examples/.
7. Add Makefile validation.

Acceptance criteria:

• make validate passes.
• Repo states lab-only boundary and governance handoff.
• Manifest can feed Holmes, Sherlock Search, model-router, model-governance-ledger, guardrail-fabric, and SourceOS carry refs.
• No large model weights, datasets, secrets, or credentials are committed.

Boundary:
One PR only. Do not touch unrelated repos.

[mdheller]

@codex Please take this imagelab issue. Implement the scoped lab maturity spine and vision service manifest in the issue body, then open one PR with validation evidence.

[mdheller]

@copilot Please take this imagelab issue if available. Implement the scoped lab maturity spine and vision service manifest, then open one PR with validation evidence.

[chatgpt-codex-connector]

Summary

• Implemented the scoped lab maturity spine and repository boundary documentation in README.md, including explicit lab-only constraints and governance handoff targets (Holmes, Sherlock Search, model-router, model-governance-ledger, guardrail-fabric, SourceOS carry refs). README.mdL1-L58
• Added repo.maturity.yaml using repo-maturity.v1 semantics with in-scope/out-of-scope definitions, staged maturity spine, and control requirements for secrets/weights/dataset handling and validation reproducibility. repo.maturity.yamlL1-L46
• Added lab.manifest.json with the required vision capability set (classification, captioning, object/region detection, chart/diagram extraction, image evidence extraction) plus multimodal handoff metadata to Holmes and integration targets. lab.manifest.jsonL1-L51
• Added service-manifest/functional-service.v1.json to define a governed functional service surface, I/O contracts, and governance feed/control hooks for downstream systems. service-manifest/functional-service.v1.jsonL1-L51
• Added required scaffold documentation for lab data/evals/adapters/training/examples:
  • datasets/README.md
  • evals/README.md
  • adapters/README.md
  • training-runs/README.md
  • examples/README.md. datasets/README.mdL1-L8 evals/README.mdL1-L8 adapters/README.mdL1-L5 training-runs/README.mdL1-L11 examples/README.mdL1-L3
• Fixed and normalized Makefile target formatting (including validate) and expanded tools/validate.sh to assert the new required directories/files and JSON manifest parse validity. MakefileL1-L19 tools/validate.shL1-L40

Testing

• ✅ make validate (passes; includes JSON parse checks and structural checks via tools/validate.sh). tools/validate.shL7-L40

Also completed follow-up actions requested: committed changes (be1ea44) and opened a PR via the make_pr tool with validation evidence included.

View task →