@gobing-ai/spur 0.3.49 → 0.3.50

package/package.json

@@ -1,6 +1,6 @@
 {
     "name": "@gobing-ai/spur",
-    "version": "0.3.49",
+    "version": "0.3.50",
     "description": "Spur CLI — local-first harness for mainstream coding agents: constraint checking, workflow orchestration, agent health, and history analytics. Bun-native; exposes the `spur` command.",
     "keywords": [
         "spur",
@@ -53,14 +53,14 @@
     },
     "devDependencies": {
         "@commander-js/extra-typings": "^14.0.0",
-        "@gobing-ai/ts-db": "^0.4.35",
-        "@gobing-ai/ts-ai-runner": "^0.4.35",
-        "@gobing-ai/ts-dual-workflow-engine": "^0.4.35",
-        "@gobing-ai/ts-infra": "^0.4.35",
-        "@gobing-ai/ts-llm-jsonl-importer": "^0.4.31",
-        "@gobing-ai/ts-rule-engine": "^0.4.35",
-        "@gobing-ai/ts-runtime": "^0.4.35",
-        "@gobing-ai/ts-utils": "^0.4.35",
+        "@gobing-ai/ts-db": "^0.4.38",
+        "@gobing-ai/ts-ai-runner": "^0.4.38",
+        "@gobing-ai/ts-dual-workflow-engine": "^0.4.38",
+        "@gobing-ai/ts-infra": "^0.4.38",
+        "@gobing-ai/ts-llm-jsonl-importer": "^0.4.38",
+        "@gobing-ai/ts-rule-engine": "^0.4.38",
+        "@gobing-ai/ts-runtime": "^0.4.38",
+        "@gobing-ai/ts-utils": "^0.4.38",
         "@types/bun": "1.3.14",
         "@types/figlet": "^1.7.0",
         "@types/node-notifier": "8.0.5",

package/plugins/sp/plugin.json

@@ -1,6 +1,6 @@
 {
     "name": "sp",
-    "version": "0.3.49",
+    "version": "0.3.50",
     "description": "Spur — a local-first harness engineering toolkit that wraps mainstream coding agents with constraint checking, workflow orchestration, and history analytics.",
     "extensions": {
         "pi": ["./hooks/pi/guard-extension.ts"]

package/plugins/sp/skills/code-verification/SKILL.md

@@ -127,6 +127,20 @@ heading/comment unrelated to the requirement fails the row to UNMET and surfaces
 a finding (severity >= P2). This closes the gap where a verify run certified a task `done` citing
 `evidence:134` that was actually a sibling ticket's telemetry text (0299 R1, from the 0282 re-audit).
 
+**External evidence (task 0584 / ADR-062) — legal citation form.** Evidence that lives OUTSIDE this
+repository has a frozen non-anchor form: a named origin plus a backticked path with the line number
+**outside** the backticks.
+
+```
+Evidence: @gobing-ai/ts-llm-jsonl-importer `src/mappers.ts` line 481 — omp call_id write
+```
+
+Use this form ONLY when the evidence genuinely lives outside `spur`'s working tree (a `@gobing-ai/ts-*`
+source under `~/xprojects/ts-libs`, or a gitignored `.spur/run/**` artifact). `checkLineAnchors`
+classifies it as external and never raises `L4.stale-line-anchor` for it (R1). Do NOT use it for a file
+that lives in this repo — in-repo evidence MUST use the repo-relative backtick form
+`` `path:line` `` / `` `path:start-end` ``, and citing it in the external form still reports (R2).
+
 ### Step 5 — Acceptance Criteria guard
 
 If the task has a non-empty Acceptance Criteria section, evaluate every checklist item and every

package/plugins/sp/skills/spur-cli/references/tasks.md

@@ -46,6 +46,7 @@ re-reading or re-tokenizing the task.
 | `list` | List tasks, filtered | `--status <s>` `--phase <p>` `--parent <wbs>` `--feature <id>` `--folder` `--json` |
 | `refresh` | Re-scan the corpus and report counts (**`kanban.md` retired** — web Task Kanban is SSOT) | `--folder` `--json` |
 | `migrate` | One-time A17 corpus normalization pass | `--dry-run` `--folder` `--json` |
+| `migrate-anchors` | Qualify in-repo evidence anchors to repo-relative paths (0583 R1–R3) | `--dry-run` `--json` |
 | `refresh-roster <wbs>` | Regenerate a parent task's sub-task roster block in `## Plan` | `--folder` `--json` |
 | `batch-create` | Create many tasks from a validated JSON array | `--file <path>` `--folder` `--json` |
 | `record <wbs>` | Write Testing/Review from a verify verdict; optional Solution + transition | `--verdict-file <path>` `--solution-from-diff` `--transition <status>` `--folder` `--json` |

package/plugins/sp/skills/spur-dev/references/ac-style-guide.md

@@ -103,7 +103,7 @@ omission is now reported as an `ac-row-dropped` check naming the row and the unr
 `rowMatchesScenario` accepts any of these as naming the feature scenario `R3 — Foo`:
 
 | Form | Example |
-|------|---------|
+| ------ | --------- |
 | Exact title | `R3 — Foo` |
 | Bare title (R-prefix dropped) | `Foo` |
 | `Scenario:` prefix | `Scenario: R3 — Foo` |
@@ -201,7 +201,7 @@ regenerates the `## Tasks` block.
 Each resolved decision from a grilling interview (Phase 1) becomes one or more Gherkin scenarios:
 
 | Decision-tree element | Becomes |
-|-----------------------|---------|
+| ----------------------- | --------- |
 | A **locked decision** (a capability the feature commits to) | A `@core` scenario — the must-ship behavior it enables |
 | The decision's **observable outcome** (why it was chosen) | The scenario's `Then` — assert the observable, not the mechanism |
 | A decision's **preconditions / constraints** | The scenario's `Given` |
@@ -211,3 +211,19 @@ Each resolved decision from a grilling interview (Phase 1) becomes one or more G
 
 Number scenarios `R1, R2, …` sequentially, stable forever (the title is the traceability identity
 key). Use the Gherkin template at `.spur/templates/bdd/gherkin.md`.
+
+## AC altitude (task 0584 / ADR-062)
+
+A task must declare, via the `ac_altitude` frontmatter field, whether its AC scenarios are the
+feature's ship contract or a finer-grained local contract. The field is the **only** input — it is
+never inferred from `template`, `status`, or whether the AC uses Gherkin (R4).
+
+| `ac_altitude` | Meaning | DD-09 subset rule |
+| --- | --- | --- |
+| `graduating` (or absent) | Task scenarios are ship-contract criteria that graduate the feature's AC | Enforced — every task scenario must match a linked-feature scenario by normalized title. Drifted titles still report (R5). |
+| `task-local` | Task criteria sit at a finer altitude than the feature's ship contract (fix-task regression criteria, per-defect cases) | Skipped — no uncovered-scenario findings regardless of title drift (R3) |
+
+Use `task-local` for a fix/refactor task whose regression criteria are not the feature's ship
+contract. Keep a `graduating` task's scenario titles identical to the feature's so DD-09 stays
+satisfied. Absent-altitude is `graduating` — set `task-local` explicitly only where the subset rule
+truly does not apply (do not silently default new tasks to it).

package/plugins/sp/skills/spur-dev/references/cross-cutting.md

@@ -24,7 +24,6 @@ not redefine it.
 > does not restate the contract. The value table below is authoritative; parity with it is
 > enforced by `validate-flag-contracts.ts` (C3a/C3b).
 
-
 ### The one rule
 
 > **`--agent <value>` names *who* does the model-bearing work. The execution surface is derived from
@@ -49,7 +48,7 @@ selects the zero-dispatch carve-out below.
 > fallback.**
 
 | Value | Who does the work | Derived surface |
-|---|---|---|
+| --- | --- | --- |
 | `(omitted)` | The agent running this session | Host session — host-controlled; eligible model stages may use a native subagent (0508) |
 | `inline` | The agent running this session | Host session — hard guarantee: zero dispatch, never a subprocess, never a workflow hop; headless surfaces reject `inline` (exit 2, stable special error) |
 | `auto` | The role the caller declared — this command's `role:` frontmatter or the workflow step's `role:` (Layer 1, `plugins/sp/references/roles.md`); with nothing declared, `agent.default`'s role (0542) | Subprocess — a tier-resolved executor pins a specific agent/model, which the host session cannot supply |
@@ -219,7 +218,7 @@ stage is required, select the subprocess path (`--agent auto` or `--agent <name>
 Never edit a task or feature file directly. Every mutation goes through:
 
 | Intent | CLI verb |
-|--------|----------|
+| -------- | ---------- |
 | Create a task | `spur task create` |
 | Change status | `spur task update <wbs> <status>` |
 | Edit a section | `spur task update <wbs> --section <name> --from-file <path>` |
@@ -329,6 +328,15 @@ loop?**
   operator, and the *answer* is recorded where the decision belongs — the feature body, an ADR
   (`docs/00_ADR.md`), or the design doc. A task may then be created for the work the answer implies.
 
+**Authoring-contract decisions (task 0584 / ADR-062).** Two fields a task author must get right:
+
+- **External evidence** — evidence that lives outside `spur`'s working tree uses the frozen non-anchor
+  form `origin`path`line N` (line OUTSIDE the backticks), e.g.
+  `@gobing-ai/ts-llm-jsonl-importer`src/mappers.ts`line 481`. It is classified external, never
+  `L4.stale-line-anchor`. In-repo evidence MUST use the repo-relative backtick form `path:line` —
+  citing an in-repo file in the external form still reports (R1/R2). See `ac-style-guide.md` and the
+  `code-verification` skill for when each applies.
+
 **Why this is a rule and not a preference.** A decision filed as a task sits in `spur task list` and
 in a feature's Tasks table looking like queued work. It gets handed to an implementing agent, which
 either stalls or invents the decision and calls it done. It also inflates task counts, which is how
@@ -432,7 +440,7 @@ line is a claim to re-verify, not evidence to forward. Re-run and paste.
 **Red Flags — an unverified claim is usually hiding behind one of these:**
 
 | Red flag | What it usually means |
-|---|---|
+| --- | --- |
 | "This should work" / "this will pass" / "probably fine" | You are predicting, not reporting. Run it and paste the result. |
 | Expressing satisfaction ("great, that's done!") before any check ran | Relief is not evidence — the check has not been run this turn. |
 | Forwarding a subagent's "success" without re-running its gate | You are trusting a claim, not verifying it. Re-run the check yourself. |
@@ -529,7 +537,7 @@ invariants that keep the pipeline set coherent as new ones are added.
 ### Pipeline phase table
 
 | Pipeline | Lifecycle phase | Entry point | Terminal states |
-|---|---|---|---|
+| --- | --- | --- | --- |
 | `idea-pipeline.yaml` | Ideation (vague idea → feature + AC + task batch) | `/sp:dev-idea` | `handoff`, `cancelled` |
 | `planning-pipeline.yaml` | Design (known slug/task → design handoff) | `/sp:dev-plan` | `handoff`, `cancelled` |
 | `task-pipeline.yaml` | Execution (one task → done) | `/sp:dev-run` | `done`, `failed` |
@@ -691,7 +699,7 @@ The signal is emitted by the `discovery` state's brainstorm dispatch and written
 determine routing:
 
 | `design` var | `needs_design` signal | Route |
-|---|---|---|
+| --- | --- | --- |
 | `skip` | (ignored) | `decompose` (skip system-design; brainstorm summary still recorded) |
 | `auto` | `true` | `system-design` -> `design-approval` -> `decompose` |
 | `auto` | `false` | `decompose` (skip system-design) |