fix(prompts): migrate remaining 4 prompts to use DB-backed tool API instead of direct write

- research-milestone.md: replace direct write with gsd_summary_save (artifact_type: RESEARCH)
- plan-slice.md: update contradictory footer to reference gsd_plan_slice tool
- run-uat.md: replace direct write with gsd_summary_save (artifact_type: ASSESSMENT)
- complete-slice.md: update footer to reference gsd_complete_slice tool

Closes #2513

Co-authored-by: glittercowboy <186001655+glittercowboy@users.noreply.github.com>
Agent-Logs-Url: https://github.com/gsd-build/gsd-2/sessions/350eb36b-b2d7-4e1a-bd2f-debd7b9b13a3

src/resources/extensions/gsd/prompts/complete-slice.md

@@ -32,6 +32,6 @@ Then:
 11. Do not run git commands — the system commits your changes and handles any merge after this unit succeeds.
 12. Update `.gsd/PROJECT.md` if it exists — refresh current state if needed.
 
-**You MUST do ALL THREE before finishing: (1) write `{{sliceSummaryPath}}`, (2) write `{{sliceUatPath}}`, (3) call `gsd_complete_slice`. The unit will not be marked complete if any of these are missing.**
+**You MUST call `gsd_complete_slice` with the slice summary and UAT content before finishing. The tool persists to both DB and disk and renders `{{sliceSummaryPath}}` and `{{sliceUatPath}}` automatically.**
 
 When done, say: "Slice {{sliceId}} complete."

src/resources/extensions/gsd/prompts/plan-slice.md

@@ -77,6 +77,6 @@ Then:
 
 The slice directory and tasks/ subdirectory already exist. Do NOT mkdir. All work stays in your working directory: `{{workingDirectory}}`.
 
-**You MUST write the file `{{outputPath}}` before finishing.**
+**You MUST call `gsd_plan_slice` to persist the planning state before finishing.**
 
 When done, say: "Slice {{sliceId}} planned."

src/resources/extensions/gsd/prompts/research-milestone.md

@@ -28,7 +28,7 @@ Then research the codebase and relevant technologies. Narrate key findings and s
 5. **Web search budget:** You have a limited budget of web searches (max ~15 per session). Use them strategically — prefer `resolve_library` / `get_library_docs` for library documentation. Do NOT repeat the same or similar queries. If a search didn't find what you need, rephrase once or move on. Target 3-5 total web searches for a typical research unit.
 6. Use the **Research** output template from the inlined context above — include only sections that have real content
 7. If `.gsd/REQUIREMENTS.md` exists, research against it. Identify which Active requirements are table stakes, likely omissions, overbuilt risks, or domain-standard behaviors the user may or may not want.
-8. Write `{{outputPath}}`
+8. Call `gsd_summary_save` with `milestone_id: {{milestoneId}}`, `artifact_type: "RESEARCH"`, and the full research markdown as `content` — the tool computes the file path and persists to both DB and disk.
 
 ## Strategic Questions to Answer
 
@@ -42,6 +42,6 @@ Then research the codebase and relevant technologies. Narrate key findings and s
 
 **Research is advisory, not auto-binding.** Surface candidate requirements clearly instead of silently expanding scope.
 
-**You MUST write the file `{{outputPath}}` before finishing.**
+**You MUST call `gsd_summary_save` with the research content before finishing.**
 
 When done, say: "Milestone {{milestoneId}} researched."

src/resources/extensions/gsd/prompts/run-uat.md

@@ -55,7 +55,7 @@ After running all checks, compute the **overall verdict**:
 - `FAIL` — one or more checks failed
 - `PARTIAL` — some checks passed, but one or more checks were skipped, inconclusive, or still require human judgment
 
-Write `{{uatResultPath}}` with:
+Call `gsd_summary_save` with `milestone_id: {{milestoneId}}`, `slice_id: {{sliceId}}`, `artifact_type: "ASSESSMENT"`, and the full UAT result markdown as `content` — the tool computes the file path and persists to both DB and disk. The content should follow this format:
 
 ```markdown
 ---
@@ -84,6 +84,6 @@ date: <ISO 8601 timestamp>
 
 ---
 
-**You MUST write `{{uatResultPath}}` before finishing.**
+**You MUST call `gsd_summary_save` with the UAT result content before finishing.**
 
 When done, say: "UAT {{sliceId}} complete."

src/resources/extensions/gsd/tests/plan-slice-prompt.test.ts

@@ -61,6 +61,18 @@ test("plan-slice prompt: DB-backed tool names survive template substitution", ()
   assert.ok(result.includes("canonical write path"), "canonical write path language should survive substitution");
 });
 
+test("plan-slice prompt: footer references gsd_plan_slice tool, not direct write", () => {
+  const result = loadPrompt("plan-slice", { ...BASE_VARS, commitInstruction: "Do not commit." });
+  assert.ok(
+    result.includes("MUST call `gsd_plan_slice`"),
+    "footer should instruct calling gsd_plan_slice tool",
+  );
+  assert.ok(
+    !result.includes("MUST write the file"),
+    "footer should not instruct direct file write",
+  );
+});
+
 test("domain-work prompts use skillActivation placeholder", () => {
   const prompts = [
     "research-milestone",
@@ -174,6 +186,34 @@ test("research-milestone prompt substitutes skillActivation", () => {
   assert.ok(!result.includes("{{skillActivation}}"));
 });
 
+test("research-milestone prompt references gsd_summary_save, not direct write", () => {
+  const result = loadPrompt("research-milestone", {
+    workingDirectory: "/tmp/test-project",
+    milestoneId: "M001",
+    milestoneTitle: "Test Milestone",
+    milestonePath: ".gsd/milestones/M001",
+    contextPath: ".gsd/milestones/M001/M001-CONTEXT.md",
+    outputPath: "/tmp/test-project/.gsd/milestones/M001/M001-RESEARCH.md",
+    inlinedContext: "Context",
+    skillDiscoveryMode: "manual",
+    skillDiscoveryInstructions: " Discover skills manually.",
+    skillActivation: "Load research skills first.",
+  });
+
+  assert.ok(
+    result.includes("gsd_summary_save"),
+    "research-milestone should reference gsd_summary_save tool",
+  );
+  assert.ok(
+    result.includes('artifact_type: "RESEARCH"'),
+    "research-milestone should specify RESEARCH artifact type",
+  );
+  assert.ok(
+    !result.includes("MUST write the file"),
+    "research-milestone should not instruct direct file write",
+  );
+});
+
 test("research-slice prompt substitutes skillActivation", () => {
   const result = loadPrompt("research-slice", {
     workingDirectory: "/tmp/test-project",

src/resources/extensions/gsd/tests/run-uat.test.ts

@@ -228,6 +228,31 @@ test('(k) run-uat prompt template', () => {
   );
 });
 
+test('(k2) run-uat prompt references gsd_summary_save, not direct write', () => {
+  const promptResult = loadPromptFromWorktree('run-uat', {
+    workingDirectory: '/tmp/test-project',
+    milestoneId: 'M001',
+    sliceId: 'S01',
+    uatPath: '.gsd/milestones/M001/slices/S01/S01-UAT.md',
+    uatResultPath: '.gsd/milestones/M001/slices/S01/S01-UAT-RESULT.md',
+    uatType: 'artifact-driven',
+    inlinedContext: '<!-- no context -->',
+  });
+
+  assert.ok(
+    promptResult.includes('gsd_summary_save'),
+    'run-uat prompt should reference gsd_summary_save tool',
+  );
+  assert.ok(
+    promptResult.includes('artifact_type: "ASSESSMENT"'),
+    'run-uat prompt should specify ASSESSMENT artifact type',
+  );
+  assert.ok(
+    !promptResult.includes('MUST write'),
+    'run-uat prompt should not instruct direct file write in footer',
+  );
+});
+
 test('(l) dispatch preconditions via resolveSliceFile', () => {
     const base = createFixtureBase();
     const uatContent = makeUatContent('artifact-driven');