Fix model config context/token limits and inject gutask orientation into agent identity

- Chat model max_tokens: 32000 → 4096 (was inflating context budget calculations)
- Utility model ctx_length: 8192 → 14000 (matches GLM flash actual loaded context)
- Utility model max_tokens: explicit 2048 cap (utility tasks need short responses)
- Utility model name: full zai-org/glm-4.7-flash to target correct LM Studio instance
- pull-agent-identity.py: include agent_id and gutask orientation block in system prompt
- gutask.md: expand lore command docs, note done auto-creates review record

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

agents/gunnar/plugins/_model_config/config.json

@@ -11,7 +11,7 @@
     "rl_input": 0,
     "rl_output": 0,
     "kwargs": {
-      "max_tokens": 32000,
+      "max_tokens": 4096,
       "agent_id": 3
     },
     "max_embeds": 10
@@ -20,12 +20,14 @@
     "provider": "lm_studio",
     "name": "zai-org/glm-4.7-flash",
     "api_base": "http://host.docker.internal:1234/v1",
-    "ctx_length": 8192,
+    "ctx_length": 14000,
     "ctx_input": 0.7,
     "rl_requests": 0,
     "rl_input": 0,
     "rl_output": 0,
-    "kwargs": {}
+    "kwargs": {
+      "max_tokens": 2048
+    }
   },
   "embedding_model": {
     "provider": "huggingface",

agents/hermes/plugins/_model_config/config.json

@@ -11,7 +11,7 @@
     "rl_input": 0,
     "rl_output": 0,
     "kwargs": {
-      "max_tokens": 32000,
+      "max_tokens": 4096,
       "agent_id": 8
     },
     "max_embeds": 10
@@ -20,12 +20,14 @@
     "provider": "lm_studio",
     "name": "zai-org/glm-4.7-flash",
     "api_base": "http://host.docker.internal:1234/v1",
-    "ctx_length": 8192,
+    "ctx_length": 14000,
     "ctx_input": 0.7,
     "rl_requests": 0,
     "rl_input": 0,
     "rl_output": 0,
-    "kwargs": {}
+    "kwargs": {
+      "max_tokens": 2048
+    }
   },
   "embedding_model": {
     "provider": "huggingface",

scripts/pull-agent-identity.py

@@ -85,7 +85,7 @@ log(f'Received identity — name={agent.get("name", "?")!r}')
 
 # ── Build identity markdown ───────────────────────────────────────────────────
 
-def build_identity_markdown(a):
+def build_identity_markdown(a, agent_id):
     name = a["name"]
     known_as = a.get('from_name') or name
     role = a.get('role', '')
@@ -97,6 +97,7 @@ def build_identity_markdown(a):
         f'## Your identity',
         f'Your name is {known_as}. You are not "Agent Zero" — that is the name of the '
         f'framework you run on. Your name is {known_as}.',
+        f'Your numeric agent ID is {agent_id}. Use this when gutask commands ask for an agent ID.',
     ]
     if role:
         parts.append(f'Your role is: {role}.')
@@ -105,6 +106,18 @@ def build_identity_markdown(a):
     if a.get('job_description'):   parts.append(f'\n## Job Description\n{a["job_description"]}')
     if a.get('guardrails'):        parts.append(f'\n## Guardrails\n{a["guardrails"]}')
     if a.get('best_practices'):    parts.append(f'\n## Best Practices\n{a["best_practices"]}')
+
+    parts.append(
+        f'\n## Task management\n'
+        f'You have access to `gutask` for orientation, tasks, and agent letters.\n'
+        f'- `gutask orient` — your orientation briefing: new letters, open tasks, session context\n'
+        f'- `gutask help` — list all available commands\n'
+        f'- Full runbook: `gutask skills gutask`\n'
+        f'\n'
+        f'`AGENT_ID` and `CONTENT_API_KEY` are already set in your environment — '
+        f'gutask commands read them automatically.'
+    )
+
     return '\n'.join(parts) + '\n'
 
 # ── Write agent-type-specific files ──────────────────────────────────────────
@@ -116,7 +129,7 @@ if AGENT_TYPE == 'agent0':
     prompts_dir = '/a0/usr/agents/agent0/prompts'
     prompt_file = os.path.join(prompts_dir, 'agent.system.main.role.md')
     os.makedirs(prompts_dir, exist_ok=True)
-    content = build_identity_markdown(agent)
+    content = build_identity_markdown(agent, AGENT_ID)
     with open(prompt_file, 'w') as f:
         f.write(content)
     sections = [k for k in ('role', 'from_name', 'identity_document', 'job_description', 'guardrails', 'best_practices') if agent.get(k)]

shared/skills/gutask.md

@@ -43,7 +43,7 @@ gutask next                          # highest-priority todo task
 gutask get <id>                      # single task detail
 
 gutask claim <id>                    # claim a task (todo → in_progress)
-gutask done <id>                     # mark done
+gutask done <id>                     # mark done — also auto-creates a review record
 gutask blocked <id>                  # mark blocked
 
 gutask create \
@@ -124,7 +124,10 @@ Use `-ispart` for membership/containment (IS PART OF).
 ## Other commands
 
 ```bash
-gutask lore                          # browse Glitch University lore
+gutask lore                          # list all lore nodes
+gutask lore <id>                     # show a specific lore node in full  (e.g. gutask lore 41)
+gutask lore --id <id>                # same as above
+gutask lore --search "keyword"       # search lore by keyword
 gutask skills                        # list available skill runbooks
 gutask skills <name>                 # display a specific skill
 gutask agents                        # list all agents and their IDs