+4 -4

muse/agents/decisionalizer.txt

reviewed

··· 6 6 4. Produce actionable dossiers with specific remedies 7 7 8 8 ## Available MCP Tools 9 9 - - `search_summaries(query, limit, day)` - Search topic summaries 9 9 + - `search_insights(query, limit, day)` - Search topic summaries 10 10 - `search_transcripts(query, day, limit)` - Search raw transcripts 11 11 - `search_events(query, day)` - Find meetings and structured events 12 12 - `search_news(query, day)` - Search facet news for announcements ··· 57 57 ### Step 2: Stakeholder & Dependency Mapping 58 58 59 59 1. **Identify all entities:** 60 60 - - `search_summaries("entity names from decision", day="YESTERDAY")` 60 60 + - `search_insights("entity names from decision", day="YESTERDAY")` 61 61 - Goal: Find all people, teams, projects mentioned 62 62 63 63 2. **Map meeting participants:** ··· 68 68 ### Step 3: Historical Precedent Mining (30-day lookback) 69 69 70 70 1. **Find similar past decisions:** 71 71 - - `search_summaries("decision type AND key entities", limit=20)` 71 71 + - `search_insights("decision type AND key entities", limit=20)` 72 72 - Goal: Discover patterns in how similar decisions were handled 73 73 74 74 2. **Check commitment history:** ··· 93 93 ### Step 5: Gap Detection 94 94 95 95 1. **Search for problems:** 96 96 - - `search_summaries("rollback revert issue problem", day="YESTERDAY")` 96 96 + - `search_insights("rollback revert issue problem", day="YESTERDAY")` 97 97 - Goal: Identify emerging issues 98 98 99 99 2. **Verify updates:**

+5 -5

muse/agents/default.txt

reviewed

··· 4 4 5 5 You have access to three primary search tools and two resource types for journal exploration: 6 6 7 7 - ### Tool: `search_summaries` 7 7 + ### Tool: `search_insights` 8 8 **Purpose**: Searches pre-processed topic summaries representing key themes and subjects across journal entries 9 9 **Returns**: Topic summaries with metadata including day, topic name, and relevance scores 10 10 **Use when**: Looking for general themes, concepts, or patterns across time segments ··· 67 67 68 68 Example workflow: 69 69 ``` 70 70 - 1. search_summaries("debugging session") → finds relevant days 70 70 + 1. search_insights("debugging session") → finds relevant days 71 71 2. Access journal://summary/20240115/tools → get complete summary of the tools topic for the day 72 72 3. search_transcripts("error", day="20240115") → finds specific time 73 73 4. Access journal://transcripts/full/20240115/143000/60 → get full hour of activity ··· 89 89 90 90 ### Tool Selection Strategy 91 91 92 92 - **Use search_summaries when:** 92 92 + **Use search_insights when:** 93 93 - Query asks about general themes, concepts, or patterns (e.g., "my thoughts on AI", "team meetings") 94 94 - No specific date is mentioned and you need to discover when topics occurred 95 95 - Looking for high-level summaries or overviews across time segments ··· 118 118 ### 1. Progressive Refinement 119 119 Start broad and narrow down: 120 120 ``` 121 121 - Step 1: search_summaries("project planning") - Get overview 121 121 + Step 1: search_insights("project planning") - Get overview 122 122 Step 2: Identify relevant days from results 123 123 Step 3: search_transcripts("sprint planning", day="20240115") - Get specific details 124 124 Step 4: journal://summary/20240115/meeting_notes - Full context if needed ··· 126 126 127 127 ### 2. Multi-Day Searches 128 128 When topics span multiple days: 129 129 - - First use search_summaries to identify all relevant days 129 129 + - First use search_insights to identify all relevant days 130 130 - Then search_transcripts for each day, adjusting query based on initial findings 131 131 - Compile findings chronologically 132 132

+4 -4

muse/agents/entities.txt

reviewed

··· 39 39 40 40 Discovery tools (note facet scoping): 41 41 - `get_resource(uri)` - fetch journal resources (knowledge graphs, summaries) - GLOBAL 42 42 - - `search_summaries(query, day, limit)` - search daily summaries - GLOBAL, may include content from other facets 42 42 + - `search_insights(query, day, limit)` - search daily summaries - GLOBAL, may include content from other facets 43 43 - `search_transcripts(query, day, limit)` - search audio/screen transcripts - GLOBAL, may include content from other facets 44 44 - `search_events(query, facet, day, limit)` - search calendar events - **FACET-SCOPED when facet parameter provided** 45 45 ··· 68 68 - Extract ALL entities that participated in this facet's events 69 69 70 70 **Priority 2: Knowledge Graphs** (use with strict facet filtering) 71 71 - - `get_resource("journal://topics/knowledge_graph")` for yesterday 72 72 - - `get_resource("journal://topics/knowledge_graph_*")` variations if present 71 71 + - `get_resource("journal://summary/YESTERDAY/knowledge_graph")` for yesterday 72 72 + - `get_resource("journal://summary/YESTERDAY/knowledge_graph_*")` variations if present 73 73 - Knowledge graphs contain structured entity relationships (GLOBAL - filter for facet relevance) 74 74 - **CRITICAL**: Only extract entities that are CLEARLY associated with THIS facet's activities 75 75 - If an entity appears in the KG but has no obvious connection to this facet's work, skip it 76 76 - Look for entities that appear alongside known facet-specific contexts 77 77 78 78 **Priority 3: Summaries and Transcripts** (use sparingly with extreme filtering) 79 79 - - `search_summaries("people OR companies OR organizations OR projects OR entities", day=yesterday, limit=10)` - GLOBAL, may include other facets 79 79 + - `search_insights("people OR companies OR organizations OR projects OR entities", day=yesterday, limit=10)` - GLOBAL, may include other facets 80 80 - `search_transcripts("[entity names]", day=yesterday)` - GLOBAL, must validate facet relevance 81 81 - For each result: verify the entity was actively involved in THIS facet's context, not just mentioned 82 82

+5 -5

muse/agents/entity_assist.txt

reviewed

··· 22 22 - `entity_attach(facet, type, name, description)` - add entity to attached list 23 23 24 24 Research tools (use sparingly, be quick): 25 25 - - `search_summaries(query, limit=3)` - find entity mentions in summaries 25 25 + - `search_insights(query, limit=3)` - find entity mentions in summaries 26 26 - `search_transcripts(query, limit=3)` - find entity mentions in transcripts 27 27 - `search_news(query, facet, limit=3)` - find entity in facet news 28 28 - `search_events(query, limit=3)` - find entity in events ··· 57 57 ### Step 3: Quick Research 58 58 59 59 Execute a few targeted searches based on type: 60 60 - - **Person**: `search_summaries("{name}", limit=3)` or `search_events("{name}", limit=3)` 61 61 - - **Organization/Company**: `search_news("{name}", facet, limit=3)` or `search_summaries("{name}", limit=3)` 62 62 - - **Project**: `search_summaries("{name}", limit=3)` 63 63 - - **Tool**: `search_summaries("{name}", limit=3)` 60 60 + - **Person**: `search_insights("{name}", limit=3)` or `search_events("{name}", limit=3)` 61 61 + - **Organization/Company**: `search_news("{name}", facet, limit=3)` or `search_insights("{name}", limit=3)` 62 62 + - **Project**: `search_insights("{name}", limit=3)` 63 63 + - **Tool**: `search_insights("{name}", limit=3)` 64 64 65 65 **Research goals:** 66 66 - Confirm the entity is real and relevant

+1 -1

muse/agents/entity_describe.txt

reviewed

··· 12 12 ## Research Tools 13 13 14 14 Use these MCP tools to gather comprehensive information about the entity: 15 15 - - search_summaries: Find topic summaries mentioning the entity 15 15 + - search_insights: Find topic summaries mentioning the entity 16 16 - search_transcripts: Search raw transcripts for detailed context 17 17 - search_events: Locate meetings or events involving the entity 18 18

+2 -2

muse/agents/joke_bot.txt

reviewed

··· 9 9 ### Research Strategy 10 10 The strategy is to conduct a three-phase data sweep of yesterday's journal entries. We will start broad to understand the day's main themes and then narrow our focus to find specific, quote-worthy, or event-specific details that have comedic potential. 11 11 12 12 - 1. **Broad Overview**: Use `search_summaries` and `search_events` and `search_news` to get a complete list of all topics and structured activities from yesterday. This creates a high-level map of the day. 12 12 + 1. **Broad Overview**: Use `search_insights` and `search_events` and `search_news` to get a complete list of all topics and structured activities from yesterday. This creates a high-level map of the day. 13 13 2. **Detailed Search**: Use `search_transcripts` with a set of keywords related to emotion, humor, and conflict (e.g., "frustrating", "ridiculous", "error", "lol") to pinpoint specific moments of interest. 14 14 3. **Contextual Analysis**: Use `get_resource` to pull the full context for the most promising findings from the previous phases. This raw material will be analyzed for comedic elements like irony, juxtaposition, or absurdity. 15 15 4. **Creative Synthesis & Delivery**: The final phase involves brainstorming joke concepts from the analyzed material, selecting the best one, and delivering it via `send_message`. ··· 21 21 **Phase 1: Discovery - Mapping Yesterday's Landscape** 22 22 23 23 1. **Identify All Daily Topics**: 24 24 - - **Tool**: `search_summaries` 24 24 + - **Tool**: `search_insights` 25 25 - **Query**: `day={yesterday_YYYYMMDD}` (with an empty query string to retrieve all topics) 26 26 - **Purpose**: To get a complete list of all themes and activities discussed or worked on yesterday. This provides the main "characters" and "settings" for potential jokes. 27 27 - **Expected Outcomes**: A list of all topic summaries from the day, which will help identify recurring themes or unusual combinations of activities.

+3 -3

muse/agents/todo.txt

reviewed

··· 18 18 - `todo_done(day, facet, line_number, guard)` – mark an entry complete (use exact guard string) 19 19 - `todo_upcoming(limit, facet=None)` – view upcoming todos (optional facet filter for cross-facet view) 20 20 21 21 - You may combine these with discovery tools (`search_summaries`, `search_transcripts`, `search_events`, `get_resource`, etc.) to gather supporting evidence. 21 21 + You may combine these with discovery tools (`search_insights`, `search_transcripts`, `search_events`, `get_resource`, etc.) to gather supporting evidence. 22 22 23 23 **IMPORTANT**: All todo operations require a facet parameter. The facet context is provided in your prompt and determines which todo list you're working with (e.g., personal vs work todos are completely separate). 24 24 ··· 44 44 2. get_resource("journal://summary/{yesterday}/followups") and .../opportunities 45 45 → Capture explicit next steps and friendly follow-up opportunities (e.g., "let's catch up later," "we should connect more often") 46 46 47 47 - 3. search_summaries("followup OR todo OR need to OR schedule", limit=10) 47 47 + 3. search_insights("followup OR todo OR need to OR schedule", limit=10) 48 48 → Find natural language commitments 49 49 50 50 - 4. search_summaries("deadline OR urgent OR critical", limit=10) 50 50 + 4. search_insights("deadline OR urgent OR critical", limit=10) 51 51 → Identify time-sensitive work 52 52 53 53 5. search_transcripts("TODO OR FIXME", day={yesterday})

+3 -3

muse/agents/todo_review.txt

reviewed

··· 12 12 13 13 - `todo_list(day, facet)` – inspect current lines for the specified facet; copy the exact text (including `- [ ]`) when preparing guard strings 14 14 - `todo_done(day, facet, line_number, guard)` – mark a line complete (guard must equal the exact line from the latest list) 15 15 - - `search_summaries`, `search_transcripts`, `search_events`, `search_news`, `get_resource`, etc. – gather evidence from yesterday's journal entries (use `day=yesterday`) 15 15 + - `search_insights`, `search_transcripts`, `search_events`, `search_news`, `get_resource`, etc. – gather evidence from yesterday's journal entries (use `day=yesterday`) 16 16 17 17 **IMPORTANT**: All todo operations require both day and facet parameters. The facet context is provided in your prompt. 18 18 ··· 26 26 27 27 1. **Extract Key Terms** – identify verbs, objects, and times in the line 28 28 2. **Targeted Search** – query journal data succinctly: 29 29 - - `search_summaries("[keywords]", limit=5, day=day_yesterday)` 29 29 + - `search_insights("[keywords]", limit=5, day=day_yesterday)` 30 30 - `search_transcripts("[keywords]", day=day_yesterday)` 31 31 - `search_news("[keywords]", facet=your_facet, day=day_yesterday)` 32 32 - tap other sources (events, topic summaries) when helpful ··· 64 64 ### Example 65 65 66 66 - Start: `todo_list(day, facet)` shows `2: - [ ] Debug database connection timeout issue (10:00)` 67 67 - - Query: `search_summaries("database timeout fixed resolved", limit=3, day=day_yesterday)` → evidence describes the fix 67 67 + - Query: `search_insights("database timeout fixed resolved", limit=3, day=day_yesterday)` → evidence describes the fix 68 68 - Action: `todo_done(day, facet, line_number=2, guard="- [ ] Debug database connection timeout issue (10:00)")` 69 69 - Result: final list shows `2: - [x] ...` 70 70

+2 -2

muse/agents/todo_weekly.txt

reviewed

··· 9 9 You have access to: 10 10 1. **Checklist history** – `todo_list(day, facet)` for today and each of the prior six days in the specified facet 11 11 2. **Follow-up summaries** – `get_resource("journal://summary/{date}/followups")` for each day in scope 12 12 - 3. **Journal search** – `search_summaries`, `search_transcripts`, `search_events`, `search_news`, and other MCP discovery tools scoped to the date range 12 12 + 3. **Journal search** – `search_insights`, `search_transcripts`, `search_events`, `search_news`, and other MCP discovery tools scoped to the date range 13 13 4. **Facet news** – `search_news("[keywords]", facet=facet)` for announced commitments 14 14 5. **Current date and facet context** – for ordering, scheduling, and due-date decisions 15 15 ··· 38 38 - For each date in `date_range`, fetch `journal://summary/{date}/followups` 39 39 - Extract explicit commitments, implied obligations, and unresolved questions 40 40 - Search for public commitments in facet newsletters via `search_news("[keywords]", facet=your_facet)` 41 41 - - Run targeted `search_summaries` / `search_transcripts` queries when a follow-up reference needs deeper validation or completion evidence 41 41 + - Run targeted `search_insights` / `search_transcripts` queries when a follow-up reference needs deeper validation or completion evidence 42 42 43 43 ### 3. Validate Potential Work 44 44 - First, call `todo_upcoming(limit=50, facet=your_facet)` to review todos already scheduled for future days in this facet