Bundle Entry: docs/spec/GCPM_BUNDLE_v0.1.md Legacy Split Doc: Prefer the bundle entrypoint for agent retrieval; this file retains detailed, topic-local semantics.

GCPM Workflow Reports v0.1

Deterministic AI-facing workflow summaries are emitted in --json under data.report for:

gcpm add -> genesis/pkg-add-report-v0.1
gcpm remove -> genesis/pkg-remove-report-v0.1
gcpm lock -> genesis/pkg-lock-report-v0.1
gcpm update -> genesis/pkg-update-report-v0.1
gcpm run -> genesis/pkg-run-report-v0.1
gcpm build -> genesis/pkg-build-report-v0.1
gcpm install -> genesis/pkg-install-report-v0.1
gcpm verify -> genesis/pkg-verify-report-v0.1
gcpm doctor -> genesis/pkg-doctor-ai-report-v0.1
gcpm env -> genesis/pkg-env-report-v0.1
gcpm publish -> genesis/pkg-publish-report-v0.1
gcpm bridge -> genesis/pkg-bridge-report-v0.1
gcpm self-optimize -> genesis/pkg-self-optimize-report-v0.1

Purpose

These reports provide machine-actionable:

what changed
why it changed
deterministic next-step options

Shared Fields

All report objects include:

schema
workflow
changed
why
fix_options (vector of deterministic options with id, command, why)

Lock Report

lock
lock_hash
locked_count
rationale_count
rationale_artifact
strict

Update Report

lock
lock_hash
updated_count
selected_count
rationale_count
rationale_artifact

Publish Report

remote
ref
policy
depth
requested_commit
published_commit
expected_old

Reports are emitted even when publish fails, so AI agents can continue with deterministic remediation planning.

Bridge Report

ecosystem
name
version
source
source_hash
commit
snapshot
provenance_root
conversion_evidence
attestation
lock
dep_name
registry

Operational Expansion

The report surface now covers dependency mutation (add/remove), execution/build (run/build), environment realization (env), and verification/remediation loops (install/verify/doctor). This keeps agents on deterministic command-level remediation paths instead of prompt-level heuristics.