equity_research

Sector / thematic research arena. Adapted from Anthropic Cowork's market-researcher cookbook — three sub-agents (sector reader, comps spreader, note writer) produce a research-note-quality deliverable on a sector or theme. Target: free-text sector or theme + optional angle. target = { sector: "AI semiconductors", angle: "valuation reset", tickers: ["NVDA","AMD","AVGO","MRVL","ARM"] } The tickers list is the peer set the comps spreader spreads. If omitted, the sector reader is asked to surface 8-15 names first. Outputs a GenericDecision envelope to maf:decisions:out (not actions:out — this is research, not trading).

manual

Phase Pipeline

research

parallel

sector_reader (7 tools)

comps_spreader (5 tools)

→

synthesis

sequential

note_writer

Ticker

Date

Configure the arena's metadata, data sources, agents and phases. Changes write to the YAML file. Restart the app to apply.

Metadata

Description One-line summary shown on the dashboard. Schedule Cron expression. Leave blank for manual-only runs. Max iterations How many replan loops are allowed before giving up. Target key Drives which outbox publishes the verdict.

Data sources

Bind data adapters to symbolic names. Agents reference these names in their sources list.

Agents

Each agent runs a ReAct loop, calls its bound sources as tools, and emits a structured signal.

Phases

Phases run in order. parallel fans out specialists, sequential runs one-by-one, debate does N-way rounds + judge.

Smart triggers

Auto-run this arena when a Redis Stream event matches a rule. Rules run through safe_eval (boolean/comparisons, abs(), payload.field). Reference →

Ticker

Start Date

End Date

Edit arena configuration (JSON). Changes write to YAML file — restart to apply.

{
  "applicable_target_types": [],
  "description": "Sector / thematic research arena. Adapted from Anthropic Cowork\u0027s\nmarket-researcher cookbook \u2014 three sub-agents (sector reader, comps\nspreader, note writer) produce a research-note-quality deliverable on\na sector or theme.\n\nTarget: free-text sector or theme + optional angle.\n  target = { sector: \"AI semiconductors\", angle: \"valuation reset\",\n             tickers: [\"NVDA\",\"AMD\",\"AVGO\",\"MRVL\",\"ARM\"] }\n\nThe tickers list is the peer set the comps spreader spreads. If omitted,\nthe sector reader is asked to surface 8-15 names first.\n\nOutputs a GenericDecision envelope to maf:decisions:out (not actions:out\n\u2014 this is research, not trading).\n",
  "llm": {
    "deep_model": "auto",
    "deep_provider": "ollama",
    "default_provider": "ollama",
    "providers": {
      "ollama": {
        "api_key": "529224e623bb4df6bd2d35ca70830b11.1OcHilLKpOBPZbD1VizrA21P",
        "base_url": "https://ollama.com/v1",
        "max_output_tokens": 8192,
        "model": "auto",
        "temperature": 0.3
      }
    },
    "quick_model": "auto",
    "quick_provider": "ollama"
  },
  "max_iterations": 1,
  "memory": {
    "chromadb_path": "./data/chromadb",
    "embedding_model": "text-embedding-004",
    "embedding_provider": "gemini",
    "instances": []
  },
  "name": "equity_research",
  "output_stream": "maf:arena:equity_research:output",
  "phases": [
    {
      "agents": [
        {
          "applicable_target_types": [],
          "extra": {},
          "llm_tier": "deep",
          "max_react_steps": 6,
          "name": "sector_reader",
          "role": "specialist",
          "sources": [
            "latest_news",
            "sector_graph",
            "filings",
            "web",
            "macro_context",
            "eodhd_news",
            "eodhd_resolve_ticker"
          ],
          "system_prompt": "",
          "system_prompt_file": "src/maf/prompts/anthropic_fs/agents/sector_reader.md",
          "tools": []
        },
        {
          "applicable_target_types": [],
          "extra": {},
          "llm_tier": "quick",
          "max_react_steps": 6,
          "name": "comps_spreader",
          "role": "specialist",
          "sources": [
            "snapshot",
            "bars_history",
            "eodhd_fundamentals",
            "eodhd_eod",
            "eodhd_earnings_calendar"
          ],
          "system_prompt": "",
          "system_prompt_file": "src/maf/prompts/anthropic_fs/agents/comps_spreader.md",
          "tools": []
        }
      ],
      "max_rounds": 1,
      "name": "research",
      "pattern": "parallel",
      "signal_extract": false,
      "transition": "synthesis"
    },
    {
      "agents": [
        {
          "applicable_target_types": [],
          "extra": {},
          "llm_tier": "deep",
          "max_react_steps": 1,
          "name": "note_writer",
          "role": "synthesis",
          "sources": [],
          "system_prompt": "",
          "system_prompt_file": "src/maf/prompts/anthropic_fs/agents/note_writer.md",
          "tools": []
        }
      ],
      "max_rounds": 1,
      "name": "synthesis",
      "pattern": "sequential",
      "signal_extract": false,
      "transition": "END"
    }
  ],
  "sources": [
    {
      "adapter": "trtools2_api",
      "applicable_target_types": [],
      "config": {
        "query_type": "snapshot"
      },
      "name": "snapshot"
    },
    {
      "adapter": "trtools2_api",
      "applicable_target_types": [],
      "config": {
        "limit": 90,
        "lookback_days": 90,
        "query_type": "bars",
        "timeframe": "1d"
      },
      "name": "bars_history"
    },
    {
      "adapter": "trtools2_news",
      "applicable_target_types": [],
      "config": {
        "count": 50
      },
      "name": "latest_news"
    },
    {
      "adapter": "fred_api",
      "applicable_target_types": [],
      "config": {},
      "name": "macro_context"
    },
    {
      "adapter": "fomo2_knowledge",
      "applicable_target_types": [],
      "config": {},
      "name": "sector_graph"
    },
    {
      "adapter": "sec_filings",
      "applicable_target_types": [],
      "config": {},
      "name": "filings"
    },
    {
      "adapter": "web_search",
      "applicable_target_types": [],
      "config": {},
      "name": "web"
    },
    {
      "adapter": "eodhd",
      "applicable_target_types": [],
      "config": {
        "tool": "get_fundamentals"
      },
      "name": "eodhd_fundamentals"
    },
    {
      "adapter": "eodhd",
      "applicable_target_types": [],
      "config": {
        "tool": "get_earnings_calendar"
      },
      "name": "eodhd_earnings_calendar"
    },
    {
      "adapter": "eodhd",
      "applicable_target_types": [],
      "config": {
        "tool": "get_eod_data"
      },
      "name": "eodhd_eod"
    },
    {
      "adapter": "eodhd",
      "applicable_target_types": [],
      "config": {
        "tool": "resolve_ticker"
      },
      "name": "eodhd_resolve_ticker"
    },
    {
      "adapter": "eodhd",
      "applicable_target_types": [],
      "config": {
        "tool": "get_news"
      },
      "name": "eodhd_news"
    }
  ],
  "target_key": "sector_id",
  "triggers": [
    {
      "action_mode": "manual",
      "cooldown_s": 1800,
      "name": "mirofish high-impact crowd-sim mentions a sector",
      "on_stream": "mirofish:sims:emitted",
      "target": {
        "sector_id": "auto",
        "tickers": "{payload.tickers}"
      },
      "when": "payload.tickers != [] and payload.tickers != None"
    }
  ]
}

Data sources bound to this arena. Agents access these via ReAct tool calls.

Source Name	Adapter	Module	Parameters	Used By
snapshot	trtools2_api	trtools2	—	comps_spreader
bars_history	trtools2_api	trtools2	—	comps_spreader
latest_news	trtools2_news	trtools2	—	sector_reader
macro_context	fred_api	web	—	sector_reader
sector_graph	fomo2_knowledge	fomo2	—	sector_reader
filings	sec_filings	web	—	sector_reader
web	web_search	web	—	sector_reader
eodhd_fundamentals	eodhd	trtools2	—	comps_spreader
eodhd_earnings_calendar	eodhd	trtools2	—	comps_spreader
eodhd_eod	eodhd	trtools2	—	comps_spreader
eodhd_resolve_ticker	eodhd	trtools2	—	sector_reader
eodhd_news	eodhd	trtools2	—	sector_reader

Decision trail — every arena run is saved for review and auditing.

Loading memory data...

equity_research

Phase Pipeline

research

synthesis

Agent Signals

Execution Trace

Source Monitoring

Agent Reports

Debates

Decisions

Target Context

Metadata

Data sources

Agents

Phases

Smart triggers

Performance Metrics

vs Buy & Hold

Equity Curve

Trades

Signal Timeline