create-agdf 0.4.9 → 0.6.0

package/README.md

@@ -92,7 +92,7 @@ Using OpenCode's plugin installer or global config for `create-agdf` can add the
 
 The generated `.agdf/control/templates/` files are reusable starting points for durable AGDF state:
 
-- `AGDF_RUN.md` for the current run dashboard
+- `runs/<run_id>/RUN_STATE.md` for each canonical current run dashboard
 - `MASTER_BACKLOG.md` for active delivery pointers
 - `templates/artefacts/` for durable UR, PRD, SD, TP and QA report artefact templates
 - `SOT_REGISTRY.md` for one source of truth per domain
@@ -107,7 +107,7 @@ npm create agdf@latest -- init
 
 This writes:
 
-- `.agdf/control/AGDF_RUN.md`
+- `.agdf/control/runs/<run_id>/RUN_STATE.md`
 - `.agdf/control/MASTER_BACKLOG.md`
 - `.agdf/control/config.json`
 - `.agdf/control/templates/artefacts/UR.md`
@@ -138,7 +138,8 @@ npx --yes create-agdf@latest doctor --json
 
 The doctor reports missing live control files, missing current gate, missing next allowed action, empty evidence, empty backlog pointer, empty source-of-truth registry, duplicate active SoT rows and invalid quality contracts. It exits non-zero only for blocking control failures.
 
-Use `gate-check` to derive the next process decision from the doctor result and `AGDF_RUN.md`:
+Use `gate-check` to derive the next process decision from the selected canonical
+`.agdf/control/runs/<run_id>/RUN_STATE.md`. Use `--run <run_id>` or `AGDF_RUN_ID` when several runs are active:
 
 ```bash
 npm create agdf@latest -- gate-check
@@ -155,17 +156,18 @@ Use `delivery-path-search` only for high-impact planning decisions with several
 npx --yes @agdf/cli@latest delivery-path-search --surface codex --json
 ```
 
-The first release uses bounded best-first Delivery Path Search, not MCTS. It is read-only and advisory: the result must be checked by canonical `gate-check`. Codex is the executable reference adapter; Claude Code, Copilot and OpenCode reuse the same skill and contracts while declaring whether enforcement is full, tool-enforced or instruction-only.
+The runtime uses bounded best-first Delivery Path Search, not MCTS. It is read-only and advisory: the result must be checked by canonical `gate-check`. Codex and Claude Code are executable, tool-enforced evaluator adapters and support opt-in `--generate-candidates`; generated proposals supplement the deterministic baseline and are deterministically validated before evaluation. Copilot and OpenCode reuse the same skill and contracts as instruction-only surfaces until conforming executable adapters are available.
 
 Requirements and boundaries:
 
-- run it only in a repository with live `.agdf/control/AGDF_RUN.md` state
+- run it only with selected canonical `.agdf/control/runs/<run_id>/RUN_STATE.md` state
 - the current control state must expose legal next actions
 - Codex CLI must be installed and authenticated for `--surface codex`
-- `--model <id>` optionally selects the Codex evaluator model
+- Claude Code CLI must be installed and authenticated for `--surface claude`
+- `--model <id>` optionally selects the Codex or Claude evaluator model
 - `--persist` writes redacted `DELIVERY_PATH_SEARCH.json` and `.md` evidence under the current scope
 - `--fixture <path>` is for deterministic contract tests, not a production evaluator
-- Claude Code, Copilot and OpenCode have shared workflow mappings but no executable native evaluator in this release
+- Copilot and OpenCode have shared workflow mappings but no executable native evaluator in this release
 - cost units are rubric values used for bounded comparison, not measured provider currency
 
 The result is either one recommendation or `no_safe_recommendation`. In both cases run canonical `gate-check` afterwards.