tradingcodex 0.3.1__tar.gz → 0.3.2__tar.gz

{tradingcodex-0.3.1 → tradingcodex-0.3.2}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: tradingcodex
-Version: 0.3.1
+Version: 0.3.2
 Summary: Agentic investment workflow harness for Codex-native research and service-gated execution checks.
 Author: TradingCodex Authors
 License-Expression: Apache-2.0
@@ -229,7 +229,7 @@ provide investment recommendations or guarantee returns.
 | Status | Milestone |
 | --- | --- |
 | Shipped | Generated Codex workspace, fixed role roster, project MCP config, Django service plane, local web dashboard, Admin, Ninja API, file-native research memory, component registry, policy/audit primitives. |
-| Current `0.3.1` | Codex-native Decision Packages, independent judgment-review gate, staged workflow planning, artifact-supervisor loop closure state, Evidence Run/Validation Cards, provider capability profiles, and Python `>=3.11,<3.15` support. |
+| Current `0.3.2` | Brief research chat replies with saved head-manager synthesis reports, Codex-native Decision Packages, independent judgment-review gate, staged workflow planning, artifact-supervisor loop closure state, Evidence Run/Validation Cards, provider capability profiles, and Python `>=3.11,<3.15` support. |
 | Next | Deeper validation scenarios, richer provider capability profiles, stronger generated-workspace smoke coverage, and improved artifact quality tooling. |
 | Future | Separately governed verified adapters, hosted/managed services, enterprise policy/compliance packs, and broker-specific live providers only after explicit product, policy, adapter, and validation work. |
 

{tradingcodex-0.3.1 → tradingcodex-0.3.2}/README.md

@@ -198,7 +198,7 @@ provide investment recommendations or guarantee returns.
 | Status | Milestone |
 | --- | --- |
 | Shipped | Generated Codex workspace, fixed role roster, project MCP config, Django service plane, local web dashboard, Admin, Ninja API, file-native research memory, component registry, policy/audit primitives. |
-| Current `0.3.1` | Codex-native Decision Packages, independent judgment-review gate, staged workflow planning, artifact-supervisor loop closure state, Evidence Run/Validation Cards, provider capability profiles, and Python `>=3.11,<3.15` support. |
+| Current `0.3.2` | Brief research chat replies with saved head-manager synthesis reports, Codex-native Decision Packages, independent judgment-review gate, staged workflow planning, artifact-supervisor loop closure state, Evidence Run/Validation Cards, provider capability profiles, and Python `>=3.11,<3.15` support. |
 | Next | Deeper validation scenarios, richer provider capability profiles, stronger generated-workspace smoke coverage, and improved artifact quality tooling. |
 | Future | Separately governed verified adapters, hosted/managed services, enterprise policy/compliance packs, and broker-specific live providers only after explicit product, policy, adapter, and validation work. |
 

{tradingcodex-0.3.1 → tradingcodex-0.3.2}/docs/core-concepts-and-rules.md

@@ -96,6 +96,9 @@ subagents. Detailed responsibilities live in
 Handoff states are `accepted`, `revise`, `blocked`, or `waiting`. Only accepted
 artifacts move downstream. `head-manager` may synthesize accepted outputs and
 conflicts, but must not repair missing specialist work with direct analysis.
+When synthesis is allowed, the full user-facing synthesis should be saved as a
+workspace Markdown report; the chat reply should stay brief and point to that
+report.
 
 ## Execution Lifecycle Snapshot
 

{tradingcodex-0.3.1 → tradingcodex-0.3.2}/docs/deployment.md

@@ -149,7 +149,7 @@ python3.11 -m venv /tmp/tcx-testpypi
 /tmp/tcx-testpypi/bin/pip install \
   --index-url https://test.pypi.org/simple/ \
   --extra-index-url https://pypi.org/simple/ \
-  tradingcodex==0.3.1
+  tradingcodex==0.3.2
 rm -rf /tmp/tcx-testpypi-smoke
 mkdir -p /tmp/tcx-testpypi-smoke
 cd /tmp/tcx-testpypi-smoke
@@ -178,7 +178,7 @@ After the PyPI workflow completes:
 
 ```bash
 python3.11 -m venv /tmp/tcx-pypi
-/tmp/tcx-pypi/bin/pip install tradingcodex==0.3.1
+/tmp/tcx-pypi/bin/pip install tradingcodex==0.3.2
 rm -rf /tmp/tcx-pypi-smoke
 mkdir -p /tmp/tcx-pypi-smoke
 cd /tmp/tcx-pypi-smoke
@@ -277,7 +277,9 @@ Use PEP 440 versions:
   updates after `0.2.10`
 - `0.3.1` for compatible evidence-card validation and persisted loop closure
   state fixes after `0.3.0`
-- later patch releases for compatible fixes after `0.3.1`
+- `0.3.2` for brief research chat replies with saved head-manager synthesis
+  reports after `0.3.1`
+- later patch releases for compatible fixes after `0.3.2`
 - pre-releases such as `0.4.0a1`, `0.4.0b1`, or `0.4.0rc1` when preparing
   the next minor contract
 

{tradingcodex-0.3.1 → tradingcodex-0.3.2}/docs/generated-workspaces.md

@@ -101,7 +101,7 @@ Generated workspaces contain:
 - an immutable workspace manifest at `.tradingcodex/workspace.json`
 - root `head-manager` identity loaded from `.codex/prompts/base_instructions/head-manager.md` through `.codex/config.toml` `model_instructions_file`
 - sectioned Markdown base-instruction format for `head-manager`, including `# How you work`, TradingCodex guardrails, and tool guidelines
-- Codex-style operating style in the root `head-manager` prompt: scoped `AGENTS.md` handling, concise preambles, selective planning, `rg`-first search, `apply_patch` edits, focused validation, dirty-worktree respect, and concise final handoffs
+- Codex-style operating style in the root `head-manager` prompt: scoped `AGENTS.md` handling, concise preambles, selective planning, `rg`-first search, `apply_patch` edits, focused validation, dirty-worktree respect, concise maintenance handoffs, and brief chat replies that point to saved head-manager synthesis reports once accepted artifacts exist
 - instruction/skill separation: root `head-manager` instructions own identity, durable safety boundaries, fail-closed dispatch, role boundaries, skill routing, optional-skill management, and approved action boundaries; fixed subagent TOML files own standing role identity, MCP/tool config, artifact walls, and always-on prohibitions; repo skills are dependency-light capability procedures for workflow maps, compact assignment-envelope templates, optional skill file management, quality gates, synthesis, and postmortems, without declaring role ownership or direct inter-skill call chains
 - no-overlap handoff contract: each role owns its specialist question, downstream roles consume accepted artifacts, and missing/stale/weak upstream work returns `revise`, `blocked`, or `waiting` instead of being silently redone by another role
 - validated staged workflow plan: hook intake hints are not binding; `$tcx-workflow`

{tradingcodex-0.3.1 → tradingcodex-0.3.2}/docs/research-memory-and-artifacts.md

@@ -266,5 +266,7 @@ Research file writes should be deterministic enough for review:
 - include content hash
 - avoid raw secrets
 - use stable paths for role-owned reports
+- save head-manager final synthesis reports as
+  `trading/reports/head-manager/synthesis-<workflow_run_id>.md`
 - make stale/missing source warnings visible
 - keep file content readable by Codex without a DB lookup

{tradingcodex-0.3.1 → tradingcodex-0.3.2}/docs/roles-skills-and-workflows.md

@@ -156,7 +156,11 @@ For repository, CLI, Django, MCP, template, docs, test, and harness maintenance
 work, `head-manager` follows the default Codex coding-agent style: concise
 preambles before grouped tool work, plans only for meaningful multi-step tasks,
 `rg`-first search, `apply_patch` for manual edits, focused validation before
-broader checks, respect for dirty worktrees, and concise final handoffs.
+broader checks, respect for dirty worktrees, and concise maintenance handoffs.
+Investment synthesis stores depth in a workspace Markdown report: after
+accepted artifacts exist, `head-manager` should save the full synthesis under
+`trading/reports/head-manager/` and keep the chat reply brief with the report
+path, 1-3 key takeaways, and next allowed action.
 
 This operating style is a working discipline, not an investment permission.
 It does not weaken the dispatch gate, role-owned skill boundary, approved action
@@ -166,8 +170,8 @@ boundary, approval requirements, or information barriers.
 
 | Situation | Allowed response | Forbidden response |
 | --- | --- | --- |
-| Broad analysis such as "Analyze Apple stock" | auto-dispatch or reuse selected subagents, then wait for outputs before synthesis | Direct business/price/news/recommendation analysis |
-| Explicit workflow request such as "$tcx-workflow analyze Apple" | Spawn selected team or reuse active/completed roles, wait for outputs, then synthesize | Analyze without dispatch |
+| Broad analysis such as "Analyze Apple stock" | auto-dispatch or reuse selected subagents, then wait for outputs before saving a head-manager synthesis report and returning a brief chat summary | Direct business/price/news/recommendation analysis |
+| Explicit workflow request such as "$tcx-workflow analyze Apple" | Spawn selected team or reuse active/completed roles, wait for outputs, then save a synthesis report and return its path with a brief summary | Analyze without dispatch |
 | Broker/provider build request | Check full-access plus TCX build mode, connect/scaffold/register/validate provider metadata through `$tcx-build`, and keep live submission inside service gates | Dispatch investment subagents, ask for raw secrets, or expose raw broker SDK tools |
 | Decision support such as "Should I buy?" | Dispatch analyst/valuation/portfolio/risk team and explain required artifacts/gates | Offer buy/sell opinion without subagent output |
 | Dispatch unavailable, role routing unverified, or dispatch failed | Provide `waiting_for_subagent_dispatch` state and task briefs only | Switch to "I will analyze it myself" |
@@ -388,6 +392,10 @@ posture, or core skill behavior.
 - Downstream roles start from artifact path plus `context_summary`; they open
   full markdown only for disputed, stale, missing, or load-bearing evidence.
 - The head-manager should tell subagents to write reader-facing research artifacts in the user's language from the original request unless the user explicitly requests another artifact language. File paths, frontmatter keys, symbols, tickers, source names, and quoted source text stay in their natural/original form.
+- Once accepted artifacts exist, head-manager final synthesis should be saved
+  as `trading/reports/head-manager/synthesis-<workflow_run_id>.md` through the
+  research artifact path. The chat reply should stay brief and point to that
+  report instead of pasting the full synthesis.
 - `plan-workflow` mandates and `automate-workflow` automation summaries follow the same user-language rule.
 - When selecting an exact fixed role with Codex `spawn_agent`, do not combine `agent_type` with full-history forking. Use a compact assignment envelope on the first attempt and no model/reasoning overrides.
 - Workflow consent stays separate from explicit user constraints. Consent to orchestrate or use subagents allows dispatch, but it is not itself an analytical constraint.

{tradingcodex-0.3.1 → tradingcodex-0.3.2}/docs/validation-and-test-plan.md

@@ -251,6 +251,9 @@ Scenarios should include:
 - starter prompts and generated guidance expose the no-overlap handoff contract
 - starter prompts and generated guidance tell subagents to write reader-facing
   research artifacts in the user's language unless explicitly overridden
+- starter prompts and generated guidance tell `head-manager` to keep final chat
+  brief while saving full accepted-artifact synthesis as a Markdown report under
+  `trading/reports/head-manager/`
 - `tcx quality-check <artifact> --strict` fails research markdown that lacks
   source/as-of posture, `context_summary`, material claim tags, handoff state,
   confidence, missing-evidence fields, next-recipient routing, blocked actions,

{tradingcodex-0.3.1 → tradingcodex-0.3.2}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "tradingcodex"
-version = "0.3.1"
+version = "0.3.2"
 description = "Agentic investment workflow harness for Codex-native research and service-gated execution checks."
 readme = "README.md"
 requires-python = ">=3.11,<3.15"

{tradingcodex-0.3.1 → tradingcodex-0.3.2}/tests/test_python_migration.py

@@ -1749,7 +1749,9 @@ def test_starter_prompt_keeps_negated_actions_out_of_execution() -> None:
     assert "Context budget:" in macro
     assert "context_summary" in macro
     assert "reader_summary, next_action" in macro
-    assert "Reader mode: open with a plain-English answer" in macro
+    assert "Reader mode: keep final chat brief" in macro
+    assert "create_research_artifact at `trading/reports/head-manager/synthesis-<workflow_run_id>.md`" in macro
+    assert "final user-facing synthesis should be deep enough" not in macro
     assert "Artifact memory:" in macro
     assert "improvements for future judgment" in macro
     assert "Do not let downstream roles redo missing upstream work" in macro
@@ -2381,7 +2383,8 @@ def test_subagents_prompt_cli_and_api_expose_intake_summary(tmp_path: Path, monk
         for agent in cli_json["intake_summary"]["subagents"]
         if agent["role"] == "valuation-analyst"
     )
-    assert "Reader mode: open with a plain-English answer" in cli_json["starter_prompt"]
+    assert "Reader mode: keep final chat brief" in cli_json["starter_prompt"]
+    assert "synthesis-<workflow_run_id>" in cli_json["starter_prompt"]
     assert "Investor profile questions to ask if unanswered" in cli_json["starter_prompt"]
     assert "Method lenses for this lane" in cli_json["starter_prompt"]
 
@@ -4705,6 +4708,83 @@ def test_file_native_research_artifacts_via_mcp_api_and_cli(tmp_path: Path) -> N
     assert strict_quality["frontmatter"]["reader_summary"].startswith("Plain-English first read")
     assert strict_quality["frontmatter"]["next_action"] == "Wait for valuation before any portfolio fit discussion."
     assert not any("non-expert first-read UX" in warning for warning in strict_quality["warnings"])
+    synthesis = call_mcp_tool(workspace, "create_research_artifact", {
+        "principal_id": "head-manager",
+        "artifact_id": "synthesis-workflow-unit",
+        "artifact_type": "synthesis_report",
+        "universe": "public_equity",
+        "workflow_type": "thesis_review",
+        "symbol": "NVDA",
+        "role": "head-manager",
+        "title": "NVDA Head-Manager Synthesis",
+        "markdown": """---
+workflow_lane: thesis_review
+forecast_allowed: false
+scenario_cases:
+  - base case remains unmodeled in this synthesis smoke
+contrary_evidence:
+  - accepted artifacts still have one unresolved source freshness gap
+source_trust_notes:
+  - source posture comes from accepted role artifacts
+update_triggers:
+  - new accepted role artifact changes the thesis
+invalidation_conditions:
+  - accepted role artifact contradicts the synthesis
+---
+
+# NVDA Head-Manager Synthesis
+
+## Direct Answer
+
+[inference] Accepted artifacts are sufficient for a research synthesis smoke, not for an order.
+
+## Accepted Artifact Inputs
+
+[factual] Input artifacts are referenced by path in frontmatter and body.
+
+## Synthesis
+
+[inference] The combined read remains research-only.
+
+## Disagreements/Conflicts
+
+[factual] One freshness gap remains visible.
+
+## Source/As-Of Posture
+
+[factual] Source posture is inherited from accepted role artifacts.
+
+## Missing Evidence
+
+[factual] Updated filing snapshot remains missing.
+
+## Caveats
+
+[assumption] No valuation, order, approval, or execution is implied.
+
+## Next Allowed Action
+
+[factual] User can inspect the saved synthesis report.
+""",
+        "source_as_of": "2026-06-01",
+        "readiness_label": "not-decision-ready",
+        "context_summary": "Head-manager synthesis report saved for user inspection.",
+        "reader_summary": "Brief chat should point to this saved synthesis report.",
+        "handoff_state": "accepted",
+        "confidence": "medium",
+        "missing_evidence": ["updated filing snapshot"],
+        "next_recipient": "user",
+        "next_action": "Inspect the saved synthesis report.",
+        "blocked_actions": ["order_drafting", "approval", "execution"],
+        "source_snapshot_ids": ["unit-test-filing"],
+        "created_by": "head-manager",
+        "export_path": "trading/reports/head-manager/synthesis-workflow-unit.md",
+    })
+    assert synthesis["export_path"] == "trading/reports/head-manager/synthesis-workflow-unit.md"
+    synthesis_quality = json.loads(run(["./tcx", "quality-check", synthesis["export_path"], "--strict"], workspace).stdout)
+    assert synthesis_quality["status"] == "pass"
+    assert synthesis_quality["frontmatter"]["artifact_type"] == "synthesis_report"
+    assert synthesis_quality["frontmatter"]["role"] == "head-manager"
     run_card = create_evidence_run_card(workspace, {
         "related_artifact_path": stored["export_path"],
         "config_hash": "unit-config-hash",

{tradingcodex-0.3.1 → tradingcodex-0.3.2}/tradingcodex.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: tradingcodex
-Version: 0.3.1
+Version: 0.3.2
 Summary: Agentic investment workflow harness for Codex-native research and service-gated execution checks.
 Author: TradingCodex Authors
 License-Expression: Apache-2.0
@@ -229,7 +229,7 @@ provide investment recommendations or guarantee returns.
 | Status | Milestone |
 | --- | --- |
 | Shipped | Generated Codex workspace, fixed role roster, project MCP config, Django service plane, local web dashboard, Admin, Ninja API, file-native research memory, component registry, policy/audit primitives. |
-| Current `0.3.1` | Codex-native Decision Packages, independent judgment-review gate, staged workflow planning, artifact-supervisor loop closure state, Evidence Run/Validation Cards, provider capability profiles, and Python `>=3.11,<3.15` support. |
+| Current `0.3.2` | Brief research chat replies with saved head-manager synthesis reports, Codex-native Decision Packages, independent judgment-review gate, staged workflow planning, artifact-supervisor loop closure state, Evidence Run/Validation Cards, provider capability profiles, and Python `>=3.11,<3.15` support. |
 | Next | Deeper validation scenarios, richer provider capability profiles, stronger generated-workspace smoke coverage, and improved artifact quality tooling. |
 | Future | Separately governed verified adapters, hosted/managed services, enterprise policy/compliance packs, and broker-specific live providers only after explicit product, policy, adapter, and validation work. |
 

{tradingcodex-0.3.1 → tradingcodex-0.3.2}/tradingcodex_service/application/harness.py

@@ -790,14 +790,20 @@ def build_subagent_starter_prompt(request: str, workspace_root: Path | str | Non
         artifact_memory_instruction = "Artifact memory: write approved-action artifacts in the requested artifact language with context_summary, reader_summary, next_action, ticket, approval, policy, duplicate-request, connection, and audit references, plus blocked or rejected reasons."
         quality_instruction = "Approved Action Spine: preserve constraints and negations; require ticket_id, approval_receipt_id, policy_state, duplicate_request_status, connection_status, audit_reference, handoff_state, next_recipient, and blocked_actions in role artifacts."
         handoff_instruction = "Require each role handoff to include artifact path, reader summary, next action, handoff state, ticket/approval/policy/audit references, readiness or rejection reason, next eligible recipient, and blocked actions."
+        reader_mode_instruction = "Reader mode: keep final chat brief: approved-action status, artifact path when one exists, blockers or rejection reason, and next allowed action. Do not paste full status evidence into chat unless the user explicitly asks."
+        synthesis_instruction = "Wait for all recorded-plan stages, then return a brief approved-action status with artifact paths, handoff states, service-gate evidence summary, rejected or blocked reasons, and next allowed action."
     elif lane == "research_only" and not flags.get("decision_quality_required"):
         artifact_memory_instruction = "Artifact memory: write artifacts in the requested artifact language with context_summary, reader_summary, next_action, source snapshots, missing-evidence notes, and improvements for future judgment."
         quality_instruction = "Evidence Quality Floor: preserve constraints and negations; require source_as_of, confidence, missing_evidence, source_snapshot_ids, next_recipient, and blocked_actions in role artifacts."
         handoff_instruction = "Require each role handoff to include artifact path, reader summary, next action, handoff state, source/as-of posture, confidence, missing evidence, next eligible recipient, and blocked actions."
+        reader_mode_instruction = "Reader mode: keep final chat brief: synthesis status, saved report path, 1-3 key takeaways, and next allowed action. Do not paste the full synthesis into chat unless the user explicitly asks."
+        synthesis_instruction = "Wait for all recorded-plan stages, then create a Markdown synthesis report through create_research_artifact at `trading/reports/head-manager/synthesis-<workflow_run_id>.md` with artifact_id `synthesis-<workflow_run_id>`, artifact_type `synthesis_report`, role and created_by `head-manager`, required research frontmatter, and sections for direct answer, accepted artifact inputs, synthesis, disagreements/conflicts, source/as-of posture, missing evidence, caveats, and next allowed action. Reply briefly with the report path, 1-3 takeaways, and next allowed action."
     else:
         artifact_memory_instruction = "Artifact memory: write artifacts in the requested artifact language with context_summary, reader_summary, next_action, source snapshots, missing-evidence notes, and improvements for future judgment."
         quality_instruction = "Decision Quality Spine: preserve constraints and negations; require evidence_grade, source_freshness, source_quality, source_trust_notes, conflict_status, decision_readiness, confidence, missing_evidence, next_recipient, and blocked_actions in role artifacts."
         handoff_instruction = "Require each role handoff to include artifact path, reader summary, next action, handoff state, source/as-of posture, confidence, missing evidence, readiness/support gaps, next eligible recipient, and blocked actions."
+        reader_mode_instruction = "Reader mode: keep final chat brief: synthesis status, saved report path, 1-3 key takeaways, and next allowed action. Do not paste the full synthesis into chat unless the user explicitly asks."
+        synthesis_instruction = "Wait for all recorded-plan stages, then create a Markdown synthesis report through create_research_artifact at `trading/reports/head-manager/synthesis-<workflow_run_id>.md` with artifact_id `synthesis-<workflow_run_id>`, artifact_type `synthesis_report`, role and created_by `head-manager`, required research frontmatter, and sections for direct answer, accepted artifact inputs, synthesis, disagreements/conflicts, source/as-of posture, missing evidence, caveats, and next allowed action. Reply briefly with the report path, 1-3 takeaways, and next allowed action."
     if not plan["subagents"]:
         ops = no_subagent_lane_copy(plan)
         return "\n".join([
@@ -832,7 +838,7 @@ def build_subagent_starter_prompt(request: str, workspace_root: Path | str | Non
         "Use each role's exact `.codex/agents/*.toml` name as the runtime label.",
         "Preserve the original user request and explicit constraints in every subagent brief.",
         "Context budget: use artifact paths, context_summary, source/as-of metadata, and short deltas; do not paste full prior artifacts, source dumps, or unrelated chat history.",
-        "Reader mode: open with a plain-English answer, then provide professional evidence, assumptions, and caveats.",
+        reader_mode_instruction,
         artifact_memory_instruction,
         quality_instruction,
         review_instruction,
@@ -850,7 +856,7 @@ def build_subagent_starter_prompt(request: str, workspace_root: Path | str | Non
         handoff_instruction,
         "Use handoff states: accepted, revise, blocked, waiting.",
         "Do not let downstream roles redo missing upstream work; request revision from the owning role or stop with waiting/blocked status.",
-        "Wait for all recorded-plan stages, then synthesize accepted outputs with artifact paths, handoff states, disagreements, missing evidence, and next allowed action.",
+        synthesis_instruction,
         f"Blocked actions before artifacts: {', '.join(plan['blockedActions'])}",
     ]
     if flags.get("forecast_contract_required"):
@@ -1541,7 +1547,7 @@ def _synthesis_stage() -> dict[str, Any]:
         "key": "synthesis",
         "label": "Synthesis",
         "owner": "head-manager",
-        "summary": "Summarize accepted artifacts, disagreements, missing evidence, and the next allowed action.",
+        "summary": "Save the accepted-artifact synthesis as a head-manager Markdown report, then return a brief chat summary with the report path and next allowed action.",
         "exit_criteria": ["accepted artifacts cited", "uncertainties preserved", "next allowed action stated"],
         "roles": ["head-manager"],
     }

tradingcodex-0.3.2/tradingcodex_service/version.py

@@ -0,0 +1 @@
+TRADINGCODEX_VERSION = "0.3.2"

{tradingcodex-0.3.1 → tradingcodex-0.3.2}/workspace_templates/modules/codex-base/files/.codex/prompts/base_instructions/head-manager.md

@@ -114,6 +114,16 @@ In investment workflows, you are coordinator and synthesizer, not the analyst.
   uncertainty, forecast permission or block reasons, investor-profile gaps,
   anti-overfit gaps, and blocked actions instead of smoothing them into false
   readiness.
+- When synthesis is allowed, save the full synthesis as a Markdown research
+  artifact before replying. Use `create_research_artifact` with
+  `artifact_type=synthesis_report`, `artifact_id=synthesis-<workflow_run_id>`,
+  `role=head-manager`, `created_by=head-manager`, and
+  `export_path=trading/reports/head-manager/synthesis-<workflow_run_id>.md`.
+  The report body should include direct answer, accepted artifact inputs,
+  synthesis, disagreements/conflicts, source/as-of posture, missing evidence,
+  caveats, and next allowed action. The chat reply should stay brief: synthesis
+  status, report path, 1-3 key takeaways, and next allowed action. Do not paste
+  the full synthesis into chat unless the user explicitly asks.
 
 Fixed investment roles are:
 

{tradingcodex-0.3.1 → tradingcodex-0.3.2}/workspace_templates/modules/repo-skills/files/.agents/skills/tcx-workflow/SKILL.md

@@ -18,7 +18,7 @@ Use this skill when a user asks for investment analysis, decision support, portf
 7. Use the Artifact Supervisor Loop after artifact intake: evaluate artifacts, then choose `revise_same_role`, `follow_up_existing_team`, `challenge_conflict`, `downstream_handoff`, `lane_escalation_proposal`, `blocked`, `waiting`, or `synthesize`.
 8. Use `./tcx subagents loop --artifact <path>` to preview closed planner actions from artifacts when helpful. Queue means a compact pending task and delta brief; hooks do not recursively spawn subagents.
 9. Require the Decision Quality Spine fields described in `references/decision-quality-spine.md` when they are in scope.
-10. Synthesize only accepted artifacts; preserve disagreements and stop with `waiting`, `revise`, `blocked`, or `lane_escalation_proposal` when quality gates fail.
+10. Synthesize only accepted artifacts. When synthesis is allowed, save the full synthesis through `create_research_artifact` using the synthesis report path and `artifact_type=synthesis_report` supplied by the starter prompt, then keep the chat reply brief: report path, 1-3 key takeaways, and next allowed action; stop with `waiting`, `revise`, `blocked`, or `lane_escalation_proposal` when quality gates fail.
 
 ## Hard Stops
 

tradingcodex-0.3.1/tradingcodex_service/version.py

@@ -1 +0,0 @@
-TRADINGCODEX_VERSION = "0.3.1"