feat(commands): add @gittensory chat <question> — grounded LLM Q&A via local Ollama

.gittensory.yml.example

@@ -899,6 +899,7 @@ settings:
   #   e2eTestGen: false    # Auto-generated E2E test scaffolding. Default: false.
   #   planner: false       # The @gittensory plan / issue-planner completion. Default: false.
   #   summaries: false     # AI summaries/rewrite text. Default: false.
+  #   chatQa: false        # @gittensory chat <question> grounded LLM Q&A. Ollama-ONLY (never the frontier env.AI); needs env.AI_ADVISORY set. Co-requisite: commandRateLimitPolicy: hold (defaults off). Default: false.
 
 # Maintainer AI review tuning (`.gittensory.yml` top-level `review:` block). These knobs shape the advisory AI
 # review prompt and file selection only — gate/slop/secret-scan are unaffected.

apps/gittensory-ui/public/openapi.json

@@ -9167,13 +9167,18 @@
               },
               "summaries": {
                 "type": "boolean"
+              },
+              "chatQa": {
+                "type": "boolean",
+                "description": "Opt the `@gittensory chat <question>` grounded Q&A surface (#4595) into local Ollama inference. Ollama-only: unlike the four capabilities above it NEVER falls back to the frontier env.AI when off; it declines unless this is true AND env.AI_ADVISORY is configured. Co-requisite: set `commandRateLimitPolicy` to `hold` (it defaults to `off` fleet-wide) so the tighter `commandRateLimitAiMaxPerWindow` ceiling actually throttles this cost-bearing command."
               }
             },
             "required": [
               "slop",
               "e2eTestGen",
               "planner",
-              "summaries"
+              "summaries",
+              "chatQa"
             ]
           },
           "gittensorLabel": {

apps/gittensory-ui/src/lib/command-reference.ts

@@ -12,6 +12,12 @@ export const PUBLIC_COMMAND_ENTRIES = [
     description:
       "Answer contribution-quality questions from connected cached sources with citations.",
   },
+  {
+    id: "chat",
+    title: "Gittensory grounded chat Q&A",
+    description:
+      "Answer a question in natural prose from cached decision-pack facts via local Ollama (maintainer/collaborator; read-only).",
+  },
   {
     id: "preflight",
     title: "Gittensory preflight",
@@ -55,7 +61,7 @@ export const PUBLIC_COMMAND_ENTRIES = [
 ] as const;
 
 export const PUBLIC_COMMAND_LIST =
-  "@gittensory help\n@gittensory ask\n@gittensory preflight\n@gittensory blockers\n@gittensory duplicate-check\n@gittensory miner-context\n@gittensory next-action\n@gittensory reviewability\n@gittensory repo-fit\n@gittensory packet";
+  "@gittensory help\n@gittensory ask\n@gittensory chat\n@gittensory preflight\n@gittensory blockers\n@gittensory duplicate-check\n@gittensory miner-context\n@gittensory next-action\n@gittensory reviewability\n@gittensory repo-fit\n@gittensory packet";
 
 export const MAINTAINER_COMMAND_ENTRIES = [
   {

config/examples/gittensory.full.yml

@@ -912,6 +912,7 @@ settings:
   #   e2eTestGen: false    # Auto-generated E2E test scaffolding. Default: false.
   #   planner: false       # The @gittensory plan / issue-planner completion. Default: false.
   #   summaries: false     # AI summaries/rewrite text. Default: false.
+  #   chatQa: false        # @gittensory chat <question> grounded LLM Q&A. Ollama-ONLY (never the frontier env.AI); needs env.AI_ADVISORY set. Co-requisite: commandRateLimitPolicy: hold (defaults off). Default: false.
 
 # Maintainer AI review tuning (`.gittensory.yml` top-level `review:` block). These knobs shape the advisory AI
 # review prompt and file selection only — gate/slop/secret-scan are unaffected.

packages/gittensory-engine/src/focus-manifest.ts

@@ -1959,6 +1959,7 @@ function parseSettingsOverride(value: JsonValue | undefined, warnings: string[],
     if (typeof rawRouting.e2eTestGen === "boolean") sparseRouting.e2eTestGen = validated.e2eTestGen;
     if (typeof rawRouting.planner === "boolean") sparseRouting.planner = validated.planner;
     if (typeof rawRouting.summaries === "boolean") sparseRouting.summaries = validated.summaries;
+    if (typeof rawRouting.chatQa === "boolean") sparseRouting.chatQa = validated.chatQa;
     out.advisoryAiRouting = sparseRouting;
   } else if (r.advisoryAiRouting !== undefined) {
     warnings.push(`Manifest "settings.advisoryAiRouting" must be an object; ignoring it and keeping any existing policy.`);

packages/gittensory-engine/src/review/advisory-ai-routing-config.ts

@@ -5,6 +5,7 @@ export const DEFAULT_ADVISORY_AI_ROUTING: AdvisoryAiRoutingConfig = {
   e2eTestGen: false,
   planner: false,
   summaries: false,
+  chatQa: false,
 };
 
 function normalizeField(value: unknown, field: keyof AdvisoryAiRoutingConfig, warnings: string[]): boolean {
@@ -31,5 +32,6 @@ export function normalizeAdvisoryAiRoutingConfig(input: unknown, warnings: strin
     e2eTestGen: normalizeField(record.e2eTestGen, "e2eTestGen", warnings),
     planner: normalizeField(record.planner, "planner", warnings),
     summaries: normalizeField(record.summaries, "summaries", warnings),
+    chatQa: normalizeField(record.chatQa, "chatQa", warnings),
   };
 }

packages/gittensory-engine/src/settings/command-authorization.ts

@@ -14,6 +14,12 @@ export const DEFAULT_COMMAND_AUTHORIZATION_POLICY: RepositoryCommandAuthorizatio
     "noise-report": ["maintainer", "collaborator"],
     "gate-override": ["maintainer", "collaborator"],
     plan: ["maintainer", "collaborator"],
+    // #4595: deliberately narrower than "ask"'s default (which allows confirmed_miner) -- chat is Ollama-only
+    // grounded LLM generation, a materially larger surface than ask's deterministic-only answer, so v1 starts
+    // maintainer/collaborator-only. Explicit registration here (rather than falling through to `default`) also
+    // activates the pr_author-widening guard below, so a self-hoster can't accidentally yml themselves into
+    // "anyone commenting on their own PR" without it.
+    chat: ["maintainer", "collaborator"],
     // #1960 PR control-surface verbs. "review" is deliberately widenable to confirmed_miner (same self-rerun
     // precedent already applied to review-now, #824) — a confirmed miner may re-trigger review on their own PR.
     // The rest (pause/resume/resolve/configuration/explain) are conservative maintainer/collaborator-only

packages/gittensory-engine/src/types/manifest-deps-types.ts

@@ -112,6 +112,9 @@ export type AdvisoryAiRoutingConfig = {
   e2eTestGen: boolean;
   planner: boolean;
   summaries: boolean;
+  /** Grounded `@gittensory chat <question>` LLM Q&A (#4595). Ollama-only: unlike the four fields above it NEVER
+   *  falls back to the frontier env.AI when off -- it simply declines. Default false. */
+  chatQa: boolean;
 };
 
 export type ContributorBlacklistEntry = {