@gobing-ai/spur 0.4.5 → 0.4.6

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

@@ -219,7 +219,9 @@ clean at authoring time.
 ### Which tags exempt a row from executable evidence
 
 A `MET` row is silently demoted to `PARTIAL` unless it carries `test` or `command` evidence — the
-executable-evidence rule. Five tags opt a row out, because not every scenario is behavioral:
+executable-evidence rule. Five tags opt a row out, because not every scenario is behavioral. The
+tag counts only when the task author writes it on the AC item in the task file; `spur task verdict`
+ignores (and flags) a tag that appears only in the verify answer:
 
 `[doc-only]` · `[docs-only]` · `[non-behavior]` · `[advisory]` · `[non-core]`
 

package/plugins/sp/skills/spur-dev/references/done-housekeeping.md

@@ -139,8 +139,9 @@ silently incomplete (H6 shipped at 23/48 that way, with one verdict carrying an
    parses zero AC rows while lint still reports PASS.
 3. A behavioral AC marked `MET` with a non-executable evidence type (`static-ref`,
    `manual-review`, `llm-judge`) is **downgraded to PARTIAL** by `task verdict`, making the whole
-   verdict PARTIAL. Use `test`/`command` (grep-based verification counts as `command`), or tag the
-   AC id `[non-behavior]`/`[advisory]` when executable evidence genuinely doesn't apply.
+   verdict PARTIAL. Use `test`/`command` (grep-based verification counts as `command`). Only the
+   task author can exempt an AC, by tagging the AC item in the task file `[non-behavior]`/`[advisory]`;
+   the same tag added in the verify answer alone is ignored.
 
 **Invariant:** a force-done task has a non-empty `done_reason` naming the timeout, a verdict
 artifact whose AC rows cover every declared scenario, and a green lint/test run recorded in

package/spur.js

@@ -50681,7 +50681,7 @@ class DatabaseInspector {
       name: row.name,
       type: row.type ?? "",
       notNull: row.notnull === 1,
-      pk: row.pk === 1
+      pk: row.pk > 0
     }));
   }
   resolveColumnName(table, columns, column) {
@@ -86156,6 +86156,7 @@ class AgentCoordinationService {
       id: input.id,
       name: input.name ?? input.id,
       type: input.type,
+      ...input.executor !== undefined ? { executor: input.executor } : {},
       workspace: input.workspace ?? this.ctx.cwd,
       purpose: input.purpose && input.purpose.length > 0 ? input.purpose : `${input.type} agent`,
       tags: input.tags ?? [],
@@ -90130,6 +90131,30 @@ async function readVerdictArtifact(fs, runDir, wbs) {
   const { missing: _scratchMissing, ...result } = scratch;
   return result;
 }
+function citedTestPaths(evidence) {
+  const paths = new Set;
+  for (const match of evidence.matchAll(TEST_PATH_RE)) {
+    const path = (match[1] ?? "").replace(/^\.\//, "");
+    if (path !== "" && !path.startsWith("/"))
+      paths.add(path);
+  }
+  return [...paths];
+}
+async function findUnresolvedTestEvidence(fs, projectRoot, artifact) {
+  if (artifact === undefined)
+    return [];
+  const rows = [...artifact.requirements ?? [], ...artifact.acceptanceCriteria ?? []];
+  const unresolved = [];
+  for (const row of rows) {
+    if (String(row.status).toUpperCase() !== "MET" || typeof row.evidence !== "string")
+      continue;
+    for (const path of citedTestPaths(row.evidence)) {
+      if (!await fs.exists(join31(projectRoot, path)))
+        unresolved.push(`${row.id ?? "?"}: ${path}`);
+    }
+  }
+  return unresolved;
+}
 function computeAggregate(artifact) {
   const reqs = artifact.requirements ?? [];
   const acs = artifact.acceptanceCriteria ?? [];
@@ -90211,6 +90236,21 @@ function evaluateDoneTransition(input) {
   const acs = artifact.acceptanceCriteria ?? [];
   const internallyConsistentPass = artifact.verdict !== "PASS" || reqs.length > 0 || acs.length > 0;
   const effective = internallyConsistentPass ? harshnessMax(artifact.verdict, computed) : "UNKNOWN";
+  const unresolved = input.unresolvedTestEvidence ?? [];
+  if (effective === "PASS" && unresolved.length > 0) {
+    return {
+      kind: "deny",
+      verdict: "PARTIAL",
+      message: [
+        `Cannot transition task ${wbs} to done: the PASS verdict cites test files that do not exist.`,
+        `  task:    ${taskFilePath}`,
+        `  verdict: ${verdictPath}`,
+        ...unresolved.map((u) => `  missing: ${u}`),
+        `  remediation: re-run \`/sp:dev-verify ${wbs}\` with evidence naming real tests, ` + `or override with \`spur task update ${wbs} done --force-done --reason "<why>"\`.`
+      ].join(`
+`)
+    };
+  }
   if (effective === "PASS") {
     return { kind: "allow", reason: "pass" };
   }
@@ -90232,8 +90272,10 @@ function harshnessMax(a, b) {
   const rank = { PASS: 0, UNKNOWN: 1, PARTIAL: 2, FAIL: 3 };
   return rank[a] >= rank[b] ? a : b;
 }
+var TEST_PATH_RE;
 var init_done_transition_guard = __esm(() => {
   init_verify_verdict();
+  TEST_PATH_RE = /(?:^|[\s`'"([,])((?:[\w@.-]+\/)+[\w@.-]+\.(?:test|spec)\.[cm]?[jt]sx?)\b/g;
 });
 
 // ../../packages/app/src/workflow/workflow-resolver.ts
@@ -91839,7 +91881,9 @@ function buildAcIdentityIndex(taskContent, featureContent) {
     if (!label)
       continue;
     declareIdentity(label);
-    const head = label.split(/:(?:\s|$)/)[0]?.trim() ?? "";
+    const balancedHead = /^(.+?\([^()]*\))\s*:(?:\s|$)/.exec(label)?.[1];
+    const fallback = /\([^():]*:[^()]*\)/.test(label) ? null : label.split(/:(?:\s|$)/)[0];
+    const head = (balancedHead ?? fallback ?? "").trim();
     if (head && head !== label)
       declareIdentity(head);
     const leading = label.split(/\s+/)[0] ?? "";
@@ -92177,6 +92221,21 @@ async function readFeatureAcBody(fs, featuresDir, featureId) {
     return null;
   }
 }
+async function readTaskAcceptanceCriteria(wbs, deps) {
+  try {
+    const hit = await new TaskLocator({
+      fs: deps.fs,
+      tasksDir: deps.tasksDir,
+      ...deps.foldersConfig !== undefined ? { foldersConfig: deps.foldersConfig } : {}
+    }).findByWbs(wbs);
+    if (hit === null)
+      return "";
+    const doc = MarkdownDocument.parse(await deps.fs.readFile(hit.filePath), "task");
+    return stripAcFence(doc.getSection("Acceptance Criteria") ?? "");
+  } catch {
+    return "";
+  }
+}
 async function lintVerifyAnswerForTask(wbs, answerText, deps) {
   let taskContent;
   let featureId = "";
@@ -104398,7 +104457,11 @@ class AgentService {
     const phaseSuffix = phase !== undefined ? ` for phase '${phase}'` : "";
     if (executor !== undefined) {
       if (source === "default") {
-        warnAgentDefaultExecutorOnce(selector, this.ctx.output);
+        return {
+          ok: false,
+          exitCode: 2,
+          message: agentDefaultExecutorMessage(selector, this.ctx.roles)
+        };
       }
       if (executorDisabled(executor)) {
         return {
@@ -104447,13 +104510,11 @@ class AgentService {
     }
     if (source === "default") {
       const roleNames = this.ctx.roles !== undefined ? [...this.ctx.roles.keys()].join(", ") : "";
-      const names = this.ctx.agentConfig?.executors?.map((e) => e.name) ?? [];
-      const roleList = roleNames !== "" ? `role (${roleNames})` : "a role";
-      const executorList = names.length > 0 ? `configured executor (${names.join(", ")})` : "a configured executor";
+      const roleList = roleNames !== "" ? `a role (${roleNames})` : "a role";
       return {
         ok: false,
         exitCode: 2,
-        message: `Unknown agent.default value: '${selector}'. Accepted: ${roleList} or ${executorList}.`
+        message: `Unknown agent.default value: '${selector}'. Accepted: ${roleList}, e.g. agent.default: coder.`
       };
     }
     const canonical = resolveAgentName(selector);
@@ -104765,11 +104826,9 @@ function warnBareBinaryOnce(selector, output) {
   warnedBareBinary.add(selector);
   output.error(`Warning: --agent "${selector}" is a bare coding-agent binary name, not a role or a configured executor. It keeps working during the transition; prefer a role (scribe, coder, reviewer, planner) or an executor name (config/transition-shims.json: agent-bare-binary-name).`);
 }
-function warnAgentDefaultExecutorOnce(selector, output) {
-  if (warnedAgentDefaultExecutor.has(selector))
-    return;
-  warnedAgentDefaultExecutor.add(selector);
-  output.error(`Warning: agent.default "${selector}" is a configured executor name, not a role. It keeps working during the transition; prefer a role (scribe, coder, reviewer, planner) (config/transition-shims.json: agent-default-executor).`);
+function agentDefaultExecutorMessage(selector, roles) {
+  const roleNames = roles !== undefined && roles.size > 0 ? [...roles.keys()].join(", ") : "scribe, coder, reviewer, planner";
+  return `agent.default '${selector}' names a configured executor, but agent.default takes a role (${roleNames}). ` + `Fix: set agent.default to a role such as 'coder' in .spur/config.yaml, and pin '${selector}' explicitly ` + `with --agent ${selector} where you need that executor.`;
 }
 function buildDoctorRows(results, executors, roles) {
   const executorByName = new Map((executors ?? []).map((e) => [e.name, e]));
@@ -105021,7 +105080,7 @@ function buildRoutingAttribution(result) {
     ...result.model !== undefined ? { model: result.model } : {}
   };
 }
-var PidObservingProcessExecutor, RolePropagatingProcessExecutor, DOCTOR_CACHE_REL = ".spur/run/agent-doctor.json", DOCTOR_CACHE_TTL_MS = 60000, USAGE_SNAPSHOT_STALE_MS, TRACE_SAFE_SLASH_COMMAND, TRACE_SAFE_SLASH_TOKEN, SENSITIVE_FLAG, SENSITIVE_INLINE_FLAG, warnedBareBinary, warnedAgentDefaultExecutor;
+var PidObservingProcessExecutor, RolePropagatingProcessExecutor, DOCTOR_CACHE_REL = ".spur/run/agent-doctor.json", DOCTOR_CACHE_TTL_MS = 60000, USAGE_SNAPSHOT_STALE_MS, TRACE_SAFE_SLASH_COMMAND, TRACE_SAFE_SLASH_TOKEN, SENSITIVE_FLAG, SENSITIVE_INLINE_FLAG, warnedBareBinary;
 var init_agent_service = __esm(() => {
   init_src();
   init_src2();
@@ -105078,7 +105137,6 @@ var init_agent_service = __esm(() => {
   SENSITIVE_FLAG = /^--?(?:api[-_]?key|authorization|credential|password|secret|token)$/i;
   SENSITIVE_INLINE_FLAG = /^(--?(?:api[-_]?key|authorization|credential|password|secret|token)=).+$/i;
   warnedBareBinary = new Set;
-  warnedAgentDefaultExecutor = new Set;
 });
 
 // ../../packages/app/src/services/agent-trace-service.ts
@@ -105794,8 +105852,20 @@ async function prepareBoardModules(input) {
     }
   }
   let realProjectRoot;
-  const projectRootRealPath = () => {
-    realProjectRoot ??= input.probe.realPath(input.projectRoot) ?? input.projectRoot;
+  const verifiedRealPath = (index, id, path, label) => {
+    let resolved;
+    try {
+      resolved = input.probe.realPath(path);
+    } catch {
+      resolved = undefined;
+    }
+    if (resolved === undefined) {
+      throw new BoardModuleConfigError(index, id, `${label} "${path}" could not be resolved to a real path`);
+    }
+    return resolved;
+  };
+  const projectRootRealPath = (index, id) => {
+    realProjectRoot ??= verifiedRealPath(index, id, input.projectRoot, "project root");
     return realProjectRoot;
   };
   for (const [index, declaration] of input.declarations.entries()) {
@@ -105808,22 +105878,35 @@ async function prepareBoardModules(input) {
     }
     const root = resolve17(input.projectRoot, declaration.directory);
     const entryPath = resolveContained(index, input.projectRoot, declaration.directory, declaration.entry, declaration.id);
-    const stylePaths = declaration.styles.map((style) => resolveContained(index, input.projectRoot, declaration.directory, style, declaration.id));
+    const styles = declaration.styles.map((style) => ({
+      declared: style,
+      absolute: resolveContained(index, input.projectRoot, declaration.directory, style, declaration.id)
+    }));
+    const stylePaths = styles.map((style) => style.absolute);
     if (!await input.probe.isDirectory(root)) {
       throw new BoardModuleConfigError(index, declaration.id, `module directory "${declaration.directory}" was not found`);
     }
     if (!await input.probe.isFile(entryPath)) {
       throw new BoardModuleConfigError(index, declaration.id, `entry asset "${declaration.entry}" was not found in "${declaration.directory}"`);
     }
-    for (const [styleIndex, stylePath] of stylePaths.entries()) {
-      if (!await input.probe.isFile(stylePath)) {
-        throw new BoardModuleConfigError(index, declaration.id, `style asset "${declaration.styles[styleIndex]}" was not found`);
+    for (const style of styles) {
+      if (!await input.probe.isFile(style.absolute)) {
+        throw new BoardModuleConfigError(index, declaration.id, `style asset "${style.declared}" was not found`);
       }
     }
-    const realRoot = input.probe.realPath(root);
-    const realEntry = input.probe.realPath(entryPath);
-    if (realRoot !== undefined && realEntry !== undefined && (!isInsideRoot(projectRootRealPath(), realRoot) || !isInsideRoot(realRoot, realEntry))) {
-      throw new BoardModuleConfigError(index, declaration.id, `"${declaration.directory}" resolves outside the project root or its entry resolves outside the module directory`);
+    const realProject = projectRootRealPath(index, declaration.id);
+    const realRoot = verifiedRealPath(index, declaration.id, root, "module directory");
+    if (!isInsideRoot(realProject, realRoot)) {
+      throw new BoardModuleConfigError(index, declaration.id, `"${declaration.directory}" resolves outside the project root`);
+    }
+    const requireInsideModule = (declaredPath, realPath) => {
+      if (!isInsideRoot(realRoot, realPath)) {
+        throw new BoardModuleConfigError(index, declaration.id, `"${declaredPath}" resolves outside the project root or its module directory`);
+      }
+    };
+    requireInsideModule(declaration.entry, verifiedRealPath(index, declaration.id, entryPath, "entry asset"));
+    for (const style of styles) {
+      requireInsideModule(style.declared, verifiedRealPath(index, declaration.id, style.absolute, "style asset"));
     }
     descriptors.push({
       ...descriptorBase(declaration),
@@ -114496,7 +114579,8 @@ async function transitionTaskGuarded(deps, input) {
       reason: input.reason,
       artifact: loaded.artifact,
       verdictPath: loaded.path,
-      readError: loaded.readError
+      readError: loaded.readError,
+      unresolvedTestEvidence: input.forceDone === true ? [] : await findUnresolvedTestEvidence(deps.fs, dirname31(dirname31(deps.runDir)), loaded.artifact)
     });
     if (guardOutcome.kind === "noop") {
       return { kind: "noop", wbs, status: "done", message: guardOutcome.message };
@@ -115917,11 +116001,12 @@ var init_task_service = __esm(() => {
 });
 
 // ../../packages/app/src/services/task-verdict.ts
-function deriveVerdict(answerText, taskCheckPassed) {
+function deriveVerdict(answerText, taskCheckPassed, taskAcceptanceCriteria = "") {
   const requirements = extractRequirements(answerText);
   const parsedAc = extractAcceptanceCriteria(answerText);
-  const acceptanceCriteria = applyAcceptanceCriteriaEvidenceRule(parsedAc.rows);
-  const checks = extractChecks(answerText, taskCheckPassed, requirements, acceptanceCriteria, parsedAc.dropped);
+  const exempt = taskAcExemptions(taskAcceptanceCriteria);
+  const acceptanceCriteria = applyAcceptanceCriteriaEvidenceRule(parsedAc.rows, exempt);
+  const checks = extractChecks(answerText, taskCheckPassed, requirements, acceptanceCriteria, parsedAc.dropped, exempt);
   if (requirements.length === 0) {
     return {
       verdict: "UNKNOWN",
@@ -116094,21 +116179,31 @@ function normalizeEvidenceType2(raw) {
   }
   return EVIDENCE_TYPE_PRECEDENCE2.find((candidate) => tokens.includes(candidate)) ?? null;
 }
-function applyAcceptanceCriteriaEvidenceRule(rows) {
+function taskAcExemptions(taskAcceptanceCriteria) {
+  const exempt = new Set;
+  for (const line of taskAcceptanceCriteria.split(`
+`)) {
+    if (!EXEMPTION_TAG_RE.test(line))
+      continue;
+    const item = line.replace(/^\s*(?:[-*+]|\d+[.)])\s*(?:\[[ xX]\]\s*)?/, "");
+    const key = normalizeTitle(item);
+    if (key !== "")
+      exempt.add(key);
+  }
+  return exempt;
+}
+function applyAcceptanceCriteriaEvidenceRule(rows, exempt) {
   return rows.map((row) => {
-    if (row.status === "MET" && requiresExecutableEvidence(row.id) && row.evidenceType !== "test" && row.evidenceType !== "command") {
+    if (row.status === "MET" && requiresExecutableEvidence(row.id, exempt) && row.evidenceType !== "test" && row.evidenceType !== "command") {
       return { ...row, status: "PARTIAL" };
     }
     return row;
   });
 }
-function requiresExecutableEvidence(id) {
-  const normalized = id.toLowerCase();
-  const nonCore = normalized.includes("[advisory]") || normalized.includes("[non-core]");
-  const nonBehavior = normalized.includes("[non-behavior]") || normalized.includes("[docs-only]") || normalized.includes("[doc-only]");
-  return !nonCore && !nonBehavior;
+function requiresExecutableEvidence(id, exempt) {
+  return !exempt.has(normalizeTitle(id));
 }
-function extractChecks(_text, taskCheckPassed, requirements, acceptanceCriteria, droppedAcRows = []) {
+function extractChecks(_text, taskCheckPassed, requirements, acceptanceCriteria, droppedAcRows = [], exempt = new Set) {
   const checks = [
     {
       name: "spur task check",
@@ -116161,8 +116256,17 @@ function extractChecks(_text, taskCheckPassed, requirements, acceptanceCriteria,
       evidence: `${hollowMet.length} MET row(s) carry no recorded evidence and cannot support PASS: ` + `${hollowMet.join(", ")}. Re-run verify with populated evidence per row.`
     });
   }
+  const answerOnlyTags = acceptanceCriteria.filter((row) => EXEMPTION_TAG_RE.test(row.id) && requiresExecutableEvidence(row.id, exempt)).map((row) => row.id);
+  if (answerOnlyTags.length > 0) {
+    checks.push({
+      name: "ac-exemption-not-in-task",
+      status: "warn",
+      severity: "advisory",
+      evidence: `${answerOnlyTags.length} AC row(s) carry an evidence-rule exemption tag the task file does not: ` + `${answerOnlyTags.join("; ")}. Exemptions come only from the task's Acceptance Criteria; ` + "these rows still need test or command evidence."
+    });
+  }
   if (acceptanceCriteria.length > 0) {
-    const downgraded = acceptanceCriteria.filter((row) => row.status === "PARTIAL" && requiresExecutableEvidence(row.id) && row.evidenceType !== "test" && row.evidenceType !== "command");
+    const downgraded = acceptanceCriteria.filter((row) => row.status === "PARTIAL" && requiresExecutableEvidence(row.id, exempt) && row.evidenceType !== "test" && row.evidenceType !== "command");
     checks.push({
       name: downgraded.length > 0 ? "evidence-rule-failed" : "evidence-rule-pass",
       status: downgraded.length > 0 ? "fail" : "pass",
@@ -116193,10 +116297,12 @@ function aggregateBatchVerdicts(results) {
   const summary = `${counts.PASS} PASS, ${counts.PARTIAL} PARTIAL, ${counts.FAIL} FAIL, ${notStarted.length} NOT-STARTED (excluded)`;
   return { verdict, rolledUp, notStarted, summary };
 }
-var EVIDENCE_TYPE_PRECEDENCE2;
+var EVIDENCE_TYPE_PRECEDENCE2, EXEMPTION_TAG_RE;
 var init_task_verdict = __esm(() => {
+  init_src2();
   init_verify_verdict();
   EVIDENCE_TYPE_PRECEDENCE2 = ["test", "command", "static-ref", "manual-review", "llm-judge", "n/a"];
+  EXEMPTION_TAG_RE = /\[(?:advisory|non-core|non-behavior|docs-only|doc-only)\]/i;
 });
 
 // ../../packages/app/src/services/token-ledger-service.ts
@@ -131412,7 +131518,8 @@ Recover with: ${err.recoveryCommand}` : err.message;
       return;
     }
     const taskCheckPassed = true;
-    const result = deriveVerdict(answerText, taskCheckPassed);
+    const taskAc = await readTaskAcceptanceCriteria(wbs, corpusDeps);
+    const result = deriveVerdict(answerText, taskCheckPassed, taskAc);
     const gap = await verdictScenarioKeyGap(wbs, [...result.requirements, ...result.acceptanceCriteria ?? []], corpusDeps);
     if (gap !== null) {
       writeJsonError(context.output, options, `Verdict rows key to no scenario of linked feature ${gap.featureId}: ${summarizeRowIds(gap.offendingIds)}. Key each row by ${VERDICT_SCENARIO_KEY_FORMS} (repair: /sp:dev-verify ${wbs})`);
@@ -132609,8 +132716,9 @@ ${prompt}`;
   };
 }
 function drainAgentSelector(spec, context) {
-  if (spec.executor === undefined)
-    return spec.type;
+  if (spec.executor === undefined) {
+    throw new Error(`Spec "${spec.id}" has no executor field \u2014 add \`executor: <name>\` (an agent.executors entry) to the spec under .spur/agents/ or re-materialize it, then drain again`);
+  }
   try {
     resolveExecutor(spec.executor, context.agentConfig, { isCanonicalAgent: isAgentName });
   } catch (error) {
@@ -134313,7 +134421,7 @@ import { createRequire as createRequire3 } from "module";
 var CLI_CONFIG = {
   binaryName: "spur",
   binaryLabel: "spur",
-  binaryVersion: "0.4.5",
+  binaryVersion: "0.4.6",
   configDir: ".spur",
   configFile: ".spur/config.yaml",
   databaseFile: ".spur/spur.db"
@@ -142419,14 +142527,13 @@ async function serveModuleAsset(c, modules, fs) {
   const target = resolve27(declared.root, relative);
   if (!isInsideRoot(declared.root, target))
     return refusal(403, "Forbidden path");
+  if (!(await fs.stat(target))?.isFile())
+    return refusal(404, "Not Found");
   const realRoot = resolvedOrUndefined(fs, declared.root);
-  const resolvedTarget = resolvedOrUndefined(fs, target);
-  if (realRoot !== undefined && resolvedTarget !== undefined && !isInsideRoot(realRoot, resolvedTarget)) {
+  const realTarget = resolvedOrUndefined(fs, target);
+  if (realRoot === undefined || realTarget === undefined || !isInsideRoot(realRoot, realTarget)) {
     return refusal(403, "Forbidden path");
   }
-  const realTarget = resolvedTarget ?? target;
-  if (!(await fs.stat(realTarget))?.isFile())
-    return refusal(404, "Not Found");
   const file = Bun.file(realTarget);
   const headers = new Headers({
     "content-type": file.type || "application/octet-stream",
@@ -142941,7 +143048,7 @@ await init_src3();
 import { homedir as homedir14 } from "os";
 import { basename as basename20, join as join70, relative as relative14, resolve as resolve28 } from "path";
 // ../server/package.json
-var version3 = "0.4.5";
+var version3 = "0.4.6";
 
 // ../server/src/modules/health/index.ts
 var startedAt = Date.now();