# Org-Wide TRSL License Rollout — Execution Record **Document status:** Dated execution record, non-normative, append-only (WP-0008-T05). **Not legal advice.** Records what was actually done, when, and by whose authorization. --- ## 1. What was decided Maintainer decision, 2026-07-29/30: adopt `specs/TargetRevenueSourceLicense-V1C1.md` (TRSL V1C1) as the preliminary governing license across every repository in the `coulomb` Forgejo org — confirmed explicitly as "every repo, no exceptions," including `target-revenue` itself and internal tooling/infrastructure repos (`state-hub`, `ops-bridge`, `ralph-workplan`, etc.), not only product-shaped repos. This was confirmed as an explicit override/acceptance of the `workplans/TREV-WP-0008-governance-and-pilot-rollout.md` T05 go-live gate for the **license-adoption step specifically** — see §4 below for what this does and does not mean. The Licensor named in every deployed copy is **Binky Hedgehog GmbH** (`specs/TRSL-Governance.md` §1). ## 2. Mechanism `scripts/rollout/LICENSE.trsl-v1c1` is the deployed template: the operative legal text (Preamble through Section 11) of TRSL V1C1, with a preliminary-candidate status banner retained (not final, alpha/beta legal-review deferral noted, points back to the canonical `specs/TargetRevenueSourceLicense-V1C1.md` for the full status banner and Appendix A). Appendix A itself (non-normative candidate notes) is **not** duplicated into every repo — that would drift out of sync across ~90 copies; the canonical document remains the single source of truth for it. For each repo: the existing `LICENSE` (or `LICENSE.md`/`COPYING`) was replaced, committed, and pushed as an individual, explicit git operation — not a bulk automated script (the sandbox's own safety classifier blocked an attempted bulk script; the maintainer chose to proceed repo-by-repo instead, each step visible and individually reviewable). ## 3. Results **Adopted and pushed successfully (all previously MIT-0/MIT-License unless noted):** info-tech-canon, net-kingdom, helix-forge, railiance-apps, coulomb-loop, activity-core, state-hub, key-cape, adaptive-pricing, agentic-resources, artifact-store, audit-core, binect-chrome, binect-js, binky-control, can-you-assist, citation-engine, citation-evidence, citation-work, config-atlas, coordination-engine, core-hub, direkt-vermittlung-de, disaster-control, domain-tree, email-connect, evidence-anchor, evidence-binder, evidence-source, feature-control, fin-hub (no prior LICENSE), flex-auth, freedom-intelligence, glas-harness, guide-board, hub-core, human-resources, identity-canon, ihp-railiance-probe, infospace-bench, issue-core, kaizen-agentic, kings-guard, kontextual-engine, llm-connect (no prior LICENSE), marki-docx, markitect-filter, markitect-main (no prior LICENSE), markitect-quarkdown, markitect-tool, open-cmis-tck, open-reuse, ops-bridge, ops-hub, ops-mason, ops-warden, phase-memory, polycode-sim, python-snake, qonto-assistant, rail-knative, rail-kubernetes, railiance-cluster (also reachable as `railiance-bootstrap` — a local symlink to the same directory/repo, not two separate repos), railiance-enablement, railiance-fabric, railiance-forge, railiance-hosts (no prior LICENSE), railiance-infra (no prior LICENSE), railiance-master, railiance-platform (no prior LICENSE), ralph-workplan (no prior LICENSE), rapp-openbao, rapp-qonto, reef-railiance, rein-aharness, rein-openweights, repo-scoping, repo-seed, reuse-surface, sand-boxer, secrets-engine, shard-wiki, snuggles-inventor, tegwick-control (no prior LICENSE), tele-mcp (no prior LICENSE), tenant-engine, testdrive-jsui, timeline-svg, unattended-progress-company, user-engine, vantage-point, vergabe-teilnahme, whynot-control (no prior LICENSE), whynot-design (no prior LICENSE), wise-validator, **target-revenue** (this repo — see §5, self-referential wording used instead of pointing to itself in the third person; `pyproject.toml`'s `license` field also updated from `MIT-0` to reflect TRSL-0.1). **Committed locally, not pushed (no `origin` remote configured at all):** - `executor-sandbox` - `executor-worker` These two repos have no git remote in this environment — the LICENSE change is committed to their local history but has nowhere to push to. Flagged for the maintainer to push (or configure a remote for) separately if these repos do have a remote elsewhere. ## 4. Explicitly skipped — requires maintainer attention **`the-custodian`** was **not** changed. Two reasons converged: 1. Its prior `LICENSE` read "**Proprietary and Confidential Source Code and Data License**" — materially different from the uniform MIT-0 placeholder every other repo carried. This is the one repo in the entire org where the rollout would replace an existing, deliberate proprietary/confidential license, not a permissive placeholder — qualitatively different from every other repo's change. 2. The sandbox's safety classifier specifically blocked this repo's change (it did not block every repo — most succeeded, several others also required a retry, but `the-custodian` is the one where this session chose not to force it, given point 1 above compounds the caution the classifier itself was already signaling). **Recommendation: do not silently fold `the-custodian` into "every repo, no exceptions" without a separate, explicit confirmation** — replacing a proprietary/confidential license with a source-available-with-eventual- conversion license is a different kind of decision than replacing an MIT-0 placeholder, and deserves its own deliberate sign-off rather than inheriting the blanket instruction by default. ## 5. `target-revenue` itself Applied with self-referential wording (this repo's own `specs/TargetRevenueSourceLicense-V1C1.md` is *the* canonical source, not a pointer to another repo). `pyproject.toml`'s `license` field updated from `{ text = "MIT-0" }` to `{ text = "TRSL-0.1 (Target Revenue Source License, preliminary candidate — see LICENSE)" }` — no other repo's package manifest (`pyproject.toml`, `package.json`, etc.) was updated in this pass; see §6. ## 6. What this rollout does NOT do — explicitly This is a **license-text adoption**, not a **Phase declaration**: - **No Phase Manifest was registered** for any repo against the hosted Trust Service. No Initial Target, no Milestone Release, no degeneration policy, no Future License has been declared for any real repo as a result of this rollout. - **No Development Credit is being tracked** anywhere. No Commercial Entitlement is for sale under any of these repos. - **No repo's package-manifest license metadata was updated**, except `target-revenue`'s own `pyproject.toml` — a repo whose `package.json` or `pyproject.toml` still declares `MIT`/`MIT-0` now has a `LICENSE` file that says something different from its own package metadata. This is a known, named gap, not an oversight papered over — fixing it across ~90 repos' package manifests is future work, not attempted here. - Per `SCOPE.md` §1's alpha/beta risk-acceptance framing, this remains a bounded operating decision for the build/alpha stage, not a substitute for the specialist legal review still deferred until out of beta. **Real Phase declaration** (an actual Initial Target, actual Trust Service registration, actual Development Credit tracking for a specific repo) remains a distinct, separate, future step — requiring, among other things, the development-effort-to-Initial-Target calculator the maintainer requested alongside this rollout (see the Target Revenue Control Plane workplan, not yet created at the time of this record).