Compare commits

..

1 Commits

Author SHA1 Message Date
Alexander Whitestone
1294f37b44 fix: [MIGRATION] Preserve legacy the-matrix quality work before Nexus rewrite (closes #685)
Some checks failed
CI / test (pull_request) Failing after 9s
CI / validate (pull_request) Failing after 12s
Review Approval Gate / verify-review (pull_request) Failing after 3s
2026-04-10 20:17:03 -04:00
6 changed files with 128 additions and 496 deletions

View File

@@ -1,132 +1,169 @@
# Legacy Matrix Audit
# Legacy Matrix Audit — Migration Table
Purpose:
Preserve useful work from `/Users/apayne/the-matrix` before the Nexus browser shell is rebuilt.
Preserve quality work from `/Users/apayne/the-matrix` before the Nexus browser shell is rebuilt.
Canonical rule:
- `Timmy_Foundation/the-nexus` is the only canonical 3D repo.
- `/Users/apayne/the-matrix` is legacy source material, not a parallel product.
- This document is the authoritative migration table for issue #685.
## Verified Legacy Matrix State
## Verified Legacy State
Local legacy repo:
- `/Users/apayne/the-matrix`
Local legacy repo: `/Users/apayne/the-matrix`
Observed facts:
- Vite browser app exists
- `npm test` passes with `87 passed, 0 failed`
- 23 JS modules under `js/`
- package scripts include `dev`, `build`, `preview`, and `test`
- Vite browser app, vanilla JS + Three.js 0.171.0
- 24 JS modules under `js/`
- Smoke suite: 87 passed, 0 failed
- Package scripts: dev, build, preview, test
- PWA manifest + service worker
- Vite config with code-splitting (Three.js in separate chunk)
- Quality-tier system for hardware detection
- WebSocket client with reconnection, heartbeat, mock mode
- Full avatar FPS movement + PiP camera
- Sub-world portal system with zone triggers
## Known historical Nexus snapshot
## Migration Table
Useful in-repo reference point:
- `0518a1c3ae3c1d0afeb24dea9772102f5a3d9a66`
Decision key:
- **CARRY** = transplant concepts and patterns into Nexus vNext
- **ARCHIVE** = keep as reference, do not directly transplant
- **DROP** = do not preserve unless re-justified
That snapshot still contains browser-world root files such as:
- `index.html`
- `app.js`
- `style.css`
- `package.json`
- `tests/`
### Core Modules
## Rescue Candidates
| File | Lines | Capability | Decision | Why for Nexus |
|------|-------|------------|----------|---------------|
| `js/main.js` | 180 | App bootstrap, render loop, WebGL context recovery | **CARRY** | Architectural pattern. Shows clean init/teardown lifecycle, context-loss recovery, visibility pause. Nexus needs this loop but should not copy the monolithic wiring. |
| `js/world.js` | 95 | Scene, camera, renderer, grid, lights | **CARRY** | Foundational. Quality-tier-aware renderer setup, grid floor, lighting. Nexus already has a world but should adopt the tier-aware antialiasing and pixel-ratio capping. |
| `js/config.js` | 68 | Connection config via URL params + env vars | **ARCHIVE** | Pattern reference only. Nexus config should route through Hermes harness, not Vite env vars. The URL-override pattern (ws, token, mock) is worth remembering. |
| `js/quality.js` | 90 | Hardware detection, quality tier (low/medium/high) | **CARRY** | Directly useful. DPR capping, core/memory/screen heuristics, WebGL renderer sniffing. Nexus needs this for graceful degradation on Mac/iPad. |
| `js/storage.js` | 39 | Safe localStorage with in-memory fallback | **CARRY** | Small, robust, sandbox-proof. Nexus should use this or equivalent. Prevents crashes in sandboxed iframes. |
### Carry forward into Nexus vNext
### Agent System
1. `agent-defs.js`
- agent identity definitions
- useful as seed data/model for visible entities in the world
| File | Lines | Capability | Decision | Why for Nexus |
|------|-------|------------|----------|---------------|
| `js/agent-defs.js` | 30 | Agent identity data (id, label, color, role, position) | **CARRY** | Seed data model. Nexus agents should be defined similarly — data-driven, not hardcoded in render logic. Color hex helper is trivial but useful. |
| `js/agents.js` | 523 | Agent 3D objects, movement, state, connection lines, hot-add/remove | **CARRY** | Core visual system. Shared geometries (perf), movement interpolation, wallet-health stress glow, auto-placement algorithm, connection-line pulse. All valuable. Needs integration with real agent state from Hermes. |
| `js/behaviors.js` | 413 | Autonomous agent behavior state machine | **ARCHIVE** | Pattern reference. The personality-weighted behavior selection, conversation pairing, and artifact-placement system are well-designed. But Nexus behaviors should be driven by Hermes, not a client-side simulation. Keep the architecture, drop the fake-autonomy. |
| `js/presence.js` | 139 | Agent presence HUD (online/offline, uptime, state) | **CARRY** | Valuable UX. Live "who's here" panel with uptime tickers and state indicators. Needs real backend state, not mock assumptions. |
2. `agents.js`
- agent objects, state machine, connection lines
- useful for visualizing Timmy / subagents / system processes in a world-native way
### Visitor & Interaction
3. `avatar.js`
- visitor embodiment, movement, camera handling
- strongly aligned with "training ground" and "walk the world" goals
| File | Lines | Capability | Decision | Why for Nexus |
|------|-------|------------|----------|---------------|
| `js/visitor.js` | 141 | Visitor enter/leave protocol, chat input | **CARRY** | Session lifecycle. Device detection, visibility-based leave/return, chat input wiring. Directly applicable to Nexus visitor tracking. |
| `js/avatar.js` | 360 | FPS movement, PiP dual-camera, touch input | **CARRY** | Visitor embodiment. WASD + arrow movement, first/third person swap, PiP canvas, touch joystick, right-click mouse-look. Strong work. Needs tuning for Nexus world bounds. |
| `js/interaction.js` | 296 | Raycasting, click-to-select agents, info popup | **CARRY** | Essential for any browser world. OrbitControls, pointer/tap detection, agent popup with state/role, TALK button. The popup-anchoring-to-3D-position logic is particularly well done. |
| `js/zones.js` | 161 | Proximity trigger zones (portal enter/exit, events) | **CARRY** | Spatial event system. Portal traversal, event triggers, once-only zones. Nexus portals (#672) need this exact pattern. |
4. `ui.js`
- HUD, chat surfaces, overlays
- useful if rebuilt against real harness data instead of stale fake state
### Chat & Communication
5. `websocket.js`
- browser-side live bridge patterns
- useful if retethered to Hermes-facing transport
| File | Lines | Capability | Decision | Why for Nexus |
|------|-------|------------|----------|---------------|
| `js/bark.js` | 141 | Speech bubble system with typing animation | **CARRY** | Timmy's voice in-world. Typing animation, queue, auto-dismiss, emotion tags, demo bark lines. Strong expressive presence. The demo lines ("The Tower watches. The Tower remembers.") are good seed content. |
| `js/ui.js` | 285 | Chat panel, agent list, HUD, streaming tokens | **CARRY** | Chat infrastructure. Rolling chat buffer, per-agent localStorage history, streaming token display with cursor animation, HTML escaping. Needs reconnection to Hermes chat instead of WS mock. |
| `js/transcript.js` | 183 | Conversation transcript logger, export | **ARCHIVE** | Pattern reference. The rolling buffer, structured JSON entries, TXT/JSON download, HUD badge are all solid. But transcript authority should live in Hermes, not browser localStorage. Keep the UX pattern, rebuild storage layer. |
6. `transcript.js`
- local transcript capture pattern
- useful if durable truth still routes through Hermes and browser cache remains secondary
### Visual Effects
7. `ambient.js`
- mood / atmosphere system
- directly supports wizardly presentation without changing system authority
| File | Lines | Capability | Decision | Why for Nexus |
|------|-------|------------|----------|---------------|
| `js/effects.js` | 195 | Matrix rain particles + starfield | **CARRY** | Atmospheric foundation. Quality-tier particle counts, frame-skip optimization, adaptive draw-range (FPS-budget recovery), bounding-sphere pre-compute. This is production-grade particle work. |
| `js/ambient.js` | 212 | Mood-driven atmosphere (lighting, fog, rain, stars) | **CARRY** | Scene mood engine. Smooth eased transitions between mood states (calm, focused, excited, contemplative, stressed), per-mood lighting/fog/rain/star parameters. Directly supports Nexus atmosphere. |
| `js/satflow.js` | 261 | Lightning payment particle flow | **CARRY** | Economy visualization. Bezier-arc particles, staggered travel, burst-on-arrival, pooling. If Nexus shows any payment/economy flow, this is the pattern. |
8. `satflow.js`
- visual economy / payment flow motifs
- useful if Timmy's economy/agent interactions become a real visible layer
### Economy & Scene
9. `economy.js`
- treasury / wallet panel ideas
- useful if later backed by real sovereign metrics
| File | Lines | Capability | Decision | Why for Nexus |
|------|-------|------------|----------|---------------|
| `js/economy.js` | 100 | Wallet/treasury HUD panel | **ARCHIVE** | UI pattern reference. Clean sats formatting, per-agent balance rows, health-colored dots, recent transactions. Worth rebuilding when backed by real sovereign metrics. |
| `js/scene-objects.js` | 718 | Dynamic 3D object registry, portals, sub-worlds | **CARRY** | Critical. Geometry/material factories, animation system (rotate/bob/pulse/orbit), portal visual (torus ring + glow disc + zone), sub-world load/unload, text sprites, compound groups. This is the most complex and valuable module. Nexus portals (#672) should build on this. |
10. `presence.js`
- who-is-here / online-state UI
- useful for showing human + agent + process presence in the world
### Backend Bridge
11. `interaction.js`
- clicking, inspecting, selecting world entities
- likely needed in any real browser-facing Nexus shell
| File | Lines | Capability | Decision | Why for Nexus |
|------|-------|------------|----------|---------------|
| `js/websocket.js` | 598 | WebSocket client, message dispatcher, mock mode | **ARCHIVE** | Pattern reference only. Reconnection with exponential backoff, heartbeat/zombie detection, rich message dispatch (40+ message types), streaming chat support. The architecture is sound but must be reconnected to Hermes transport, not copied wholesale. The message-type catalog is the most valuable reference artifact. |
| `js/demo.js` | ~300 | Demo autopilot (mock mode simulation) | **DROP** | Fake activity simulation. Deliberately creates the illusion of live data. Do not preserve. If Nexus needs a demo mode, build a clearly-labeled one that doesn't pretend to be real. |
12. `quality.js`
- hardware-aware quality tiering
- useful for local-first graceful degradation on Mac hardware
### Testing & Build
13. `bark.js`
- prominent speech / bark system
- strong fit for Timmy's expressive presence in-world
| File | Lines | Capability | Decision | Why for Nexus |
|------|-------|------------|----------|---------------|
| `test/smoke.mjs` | 235 | Automated browser smoke test suite | **CARRY** | Testing discipline. Module inventory check, export verification, HTML structure validation, Vite build test, bundle-size budget, PWA manifest check. Nexus should adopt this pattern (adapted for its own module structure). |
| `vite.config.js` | 53 | Build config with code splitting, SW generation | **ARCHIVE** | Build tooling reference. manualChunks for Three.js, SW precache generation plugin. Relevant if Nexus re-commits to Vite. |
| `sw.js` | ~40 | Service worker with precache | **ARCHIVE** | PWA reference. Relevant only if Nexus pursues offline-first PWA. |
| `manifest.json` | ~20 | PWA manifest | **ARCHIVE** | PWA reference. |
14. `world.js`, `effects.js`, `scene-objects.js`, `zones.js`
- broad visual foundation work
- should be mined for patterns, not blindly transplanted
### Server-Side (Python)
15. `test/smoke.mjs`
- browser smoke discipline
- should inform rebuilt validation in canonical Nexus repo
| File | Lines | Capability | Decision | Why for Nexus |
|------|-------|------------|----------|---------------|
| `server/bridge.py` | ~900 | WebSocket bridge server | **ARCHIVE** | Reference. Hermes replaces this role. Keep for protocol schema reference. |
| `server/gateway.py` | ~400 | HTTP gateway | **ARCHIVE** | Reference. |
| `server/ollama_client.py` | ~280 | Ollama integration | **ARCHIVE** | Reference. Relevant if Nexus needs local model calls. |
| `server/research.py` | ~450 | Research pipeline | **ARCHIVE** | Reference. |
| `server/webhooks.py` | ~350 | Webhook handler | **ARCHIVE** | Reference. |
| `server/test_*.py` | ~5 files | Server test suites | **ARCHIVE** | Testing patterns worth studying. |
### Archive as reference, not direct carry-forward
## Summary by Decision
- demo/autopilot assumptions that pretend fake backend activity is real
- any websocket schema that no longer matches Hermes truth
- Vite-specific plumbing that is only useful if we consciously recommit to Vite
### CARRY FORWARD (17 modules)
These modules contain patterns, algorithms, or entire implementations that should move into the Nexus browser shell:
### Deliberately drop unless re-justified
- `quality.js` — hardware detection
- `storage.js` — safe persistence
- `world.js` — scene foundation
- `agent-defs.js` — agent data model
- `agents.js` — agent visualization + movement
- `presence.js` — online presence HUD
- `visitor.js` — session lifecycle
- `avatar.js` — FPS embodiment
- `interaction.js` — click/select/raycast
- `zones.js` — spatial triggers
- `bark.js` — speech bubbles
- `ui.js` — chat/HUD
- `effects.js` — particle effects
- `ambient.js` — mood atmosphere
- `satflow.js` — payment flow particles
- `scene-objects.js` — dynamic objects + portals
- `test/smoke.mjs` — smoke test discipline
- anything that presents mock data as if it were live
- anything that duplicates a better Hermes-native telemetry path
- anything that turns the browser into the system of record
### ARCHIVE AS REFERENCE (9 modules/files)
Keep for patterns, protocol schemas, and architectural reference. Do not directly transplant:
- `config.js` — config pattern (use Hermes instead)
- `behaviors.js` — behavior architecture (use Hermes-driven state)
- `transcript.js` — transcript UX (use Hermes storage)
- `economy.js` — economy UI pattern (use real metrics)
- `websocket.js` — message protocol catalog + reconnection patterns
- `vite.config.js` — build tooling
- `sw.js`, `manifest.json` — PWA reference
- `server/*.py` — server protocol schemas
### DELIBERATELY DROP (2)
Do not preserve unless re-justified:
- `demo.js` — fake activity simulation; creates false impression of live system
- `main.js` monolithic wiring — the init pattern carries, the specific module wiring does not
## Concern Separation for Nexus vNext
When rebuilding inside `the-nexus`, keep concerns separated:
When rebuilding inside `the-nexus`, keep these concerns in separate modules:
1. World shell / rendering
- scene, camera, movement, atmosphere
2. Presence and embodiment
- avatar, agent placement, selection, bark/chat surfaces
3. Harness bridge
- websocket / API bridge from Hermes truth into browser state
4. Visualization panels
- metrics, presence, economy, portal states, transcripts
5. Validation
- smoke tests, screenshot proof, provenance checks
6. Game portal layer
- Morrowind / portal-specific interaction surfaces
1. **World shell** — scene, camera, renderer, grid, lights, fog
2. **Effects layer** — rain, stars, ambient mood transitions
3. **Agent visualization** — 3D objects, labels, connection lines, movement
4. **Visitor embodiment** — avatar, FPS controls, PiP camera
5. **Interaction layer** — raycasting, selection, zones, portal traversal
6. **Communication surface** — bark, chat panel, streaming tokens
7. **Presence & HUD** — who's-online, economy panel, transcript controls
8. **Harness bridge** — WebSocket/API transport to Hermes (NOT a copy of websocket.js)
9. **Quality & config** — hardware detection, runtime configuration
10. **Smoke tests** — automated validation
Do not collapse all of this into one giant app file again.
Do not let visual shell code become telemetry authority.

View File

@@ -29,8 +29,6 @@ from typing import Any, Callable, Optional
import websockets
from bannerlord_trace import BannerlordTraceLogger
# ═══════════════════════════════════════════════════════════════════════════
# CONFIGURATION
# ═══════════════════════════════════════════════════════════════════════════
@@ -267,13 +265,11 @@ class BannerlordHarness:
desktop_command: Optional[list[str]] = None,
steam_command: Optional[list[str]] = None,
enable_mock: bool = False,
enable_trace: bool = False,
):
self.hermes_ws_url = hermes_ws_url
self.desktop_command = desktop_command or DEFAULT_MCP_DESKTOP_COMMAND
self.steam_command = steam_command or DEFAULT_MCP_STEAM_COMMAND
self.enable_mock = enable_mock
self.enable_trace = enable_trace
# MCP clients
self.desktop_mcp: Optional[MCPClient] = None
@@ -288,9 +284,6 @@ class BannerlordHarness:
self.cycle_count = 0
self.running = False
# Session trace logger
self.trace_logger: Optional[BannerlordTraceLogger] = None
# ═══ LIFECYCLE ═══
async def start(self) -> bool:
@@ -321,15 +314,6 @@ class BannerlordHarness:
# Connect to Hermes WebSocket
await self._connect_hermes()
# Initialize trace logger if enabled
if self.enable_trace:
self.trace_logger = BannerlordTraceLogger(
harness_session_id=self.session_id,
hermes_session_id=self.session_id,
)
self.trace_logger.start_session()
log.info(f"Trace logger started: {self.trace_logger.trace_id}")
log.info("Harness initialized successfully")
return True
@@ -338,12 +322,6 @@ class BannerlordHarness:
self.running = False
log.info("Shutting down harness...")
# Finalize trace logger
if self.trace_logger:
manifest = self.trace_logger.finish_session()
log.info(f"Trace saved: {manifest.trace_file}")
log.info(f"Manifest: {self.trace_logger.manifest_file}")
if self.desktop_mcp:
self.desktop_mcp.stop()
if self.steam_mcp:
@@ -729,11 +707,6 @@ class BannerlordHarness:
self.cycle_count = iteration
log.info(f"\n--- ODA Cycle {iteration + 1}/{max_iterations} ---")
# Start trace cycle
trace_cycle = None
if self.trace_logger:
trace_cycle = self.trace_logger.begin_cycle(iteration)
# 1. OBSERVE: Capture state
log.info("[OBSERVE] Capturing game state...")
state = await self.capture_state()
@@ -742,24 +715,11 @@ class BannerlordHarness:
log.info(f" Screen: {state.visual.screen_size}")
log.info(f" Players online: {state.game_context.current_players_online}")
# Populate trace with observation data
if trace_cycle:
trace_cycle.screenshot_path = state.visual.screenshot_path or ""
trace_cycle.window_found = state.visual.window_found
trace_cycle.screen_size = list(state.visual.screen_size)
trace_cycle.mouse_position = list(state.visual.mouse_position)
trace_cycle.playtime_hours = state.game_context.playtime_hours
trace_cycle.players_online = state.game_context.current_players_online
trace_cycle.is_running = state.game_context.is_running
# 2. DECIDE: Get actions from decision function
log.info("[DECIDE] Getting actions...")
actions = decision_fn(state)
log.info(f" Decision returned {len(actions)} actions")
if trace_cycle:
trace_cycle.actions_planned = actions
# 3. ACT: Execute actions
log.info("[ACT] Executing actions...")
results = []
@@ -771,13 +731,6 @@ class BannerlordHarness:
if result.error:
log.info(f" Error: {result.error}")
if trace_cycle:
trace_cycle.actions_executed.append(result.to_dict())
# Finalize trace cycle
if trace_cycle:
self.trace_logger.finish_cycle(trace_cycle)
# Send cycle summary telemetry
await self._send_telemetry({
"type": "oda_cycle_complete",
@@ -883,18 +836,12 @@ async def main():
default=1.0,
help="Delay between iterations in seconds (default: 1.0)",
)
parser.add_argument(
"--trace",
action="store_true",
help="Enable session trace logging to ~/.timmy/traces/bannerlord/",
)
args = parser.parse_args()
# Create harness
harness = BannerlordHarness(
hermes_ws_url=args.hermes_ws,
enable_mock=args.mock,
enable_trace=args.trace,
)
try:

View File

@@ -1,234 +0,0 @@
#!/usr/bin/env python3
"""
Bannerlord Session Trace Logger — First-Replayable Training Material
Captures one Bannerlord session as a replayable trace:
- Timestamps on every cycle
- Actions executed with success/failure
- World-state evidence (screenshots, Steam stats)
- Hermes session/log ID mapping
Storage: ~/.timmy/traces/bannerlord/trace_<session_id>.jsonl
Manifest: ~/.timmy/traces/bannerlord/manifest_<session_id>.json
Each JSONL line is one ODA cycle with full context.
The manifest bundles metadata for replay/eval.
"""
from __future__ import annotations
import json
import time
import uuid
from dataclasses import dataclass, field, asdict
from datetime import datetime, timezone
from pathlib import Path
from typing import Optional
# Storage root — local-first under ~/.timmy/
DEFAULT_TRACE_DIR = Path.home() / ".timmy" / "traces" / "bannerlord"
@dataclass
class CycleTrace:
"""One ODA cycle captured in full."""
cycle_index: int
timestamp_start: str
timestamp_end: str = ""
duration_ms: int = 0
# Observe
screenshot_path: str = ""
window_found: bool = False
screen_size: list[int] = field(default_factory=lambda: [1920, 1080])
mouse_position: list[int] = field(default_factory=lambda: [0, 0])
playtime_hours: float = 0.0
players_online: int = 0
is_running: bool = False
# Decide
actions_planned: list[dict] = field(default_factory=list)
decision_note: str = ""
# Act
actions_executed: list[dict] = field(default_factory=list)
actions_succeeded: int = 0
actions_failed: int = 0
# Metadata
hermes_session_id: str = ""
hermes_log_id: str = ""
harness_session_id: str = ""
def to_dict(self) -> dict:
return asdict(self)
@dataclass
class SessionManifest:
"""Top-level metadata for a captured session trace."""
trace_id: str
harness_session_id: str
hermes_session_id: str
hermes_log_id: str
game: str = "Mount & Blade II: Bannerlord"
app_id: int = 261550
started_at: str = ""
finished_at: str = ""
total_cycles: int = 0
total_actions: int = 0
total_succeeded: int = 0
total_failed: int = 0
trace_file: str = ""
trace_dir: str = ""
replay_command: str = ""
eval_note: str = ""
def to_dict(self) -> dict:
return asdict(self)
class BannerlordTraceLogger:
"""
Captures a single Bannerlord session as a replayable trace.
Usage:
logger = BannerlordTraceLogger(hermes_session_id="abc123")
logger.start_session()
cycle = logger.begin_cycle(0)
# ... populate cycle fields ...
logger.finish_cycle(cycle)
manifest = logger.finish_session()
"""
def __init__(
self,
trace_dir: Optional[Path] = None,
harness_session_id: str = "",
hermes_session_id: str = "",
hermes_log_id: str = "",
):
self.trace_dir = trace_dir or DEFAULT_TRACE_DIR
self.trace_dir.mkdir(parents=True, exist_ok=True)
self.trace_id = f"bl_{datetime.now(timezone.utc).strftime('%Y%m%d_%H%M%S')}_{uuid.uuid4().hex[:6]}"
self.harness_session_id = harness_session_id or str(uuid.uuid4())[:8]
self.hermes_session_id = hermes_session_id
self.hermes_log_id = hermes_log_id
self.trace_file = self.trace_dir / f"trace_{self.trace_id}.jsonl"
self.manifest_file = self.trace_dir / f"manifest_{self.trace_id}.json"
self.cycles: list[CycleTrace] = []
self.started_at: str = ""
self.finished_at: str = ""
def start_session(self) -> str:
"""Begin a trace session. Returns trace_id."""
self.started_at = datetime.now(timezone.utc).isoformat()
return self.trace_id
def begin_cycle(self, cycle_index: int) -> CycleTrace:
"""Start recording one ODA cycle."""
cycle = CycleTrace(
cycle_index=cycle_index,
timestamp_start=datetime.now(timezone.utc).isoformat(),
harness_session_id=self.harness_session_id,
hermes_session_id=self.hermes_session_id,
hermes_log_id=self.hermes_log_id,
)
return cycle
def finish_cycle(self, cycle: CycleTrace) -> None:
"""Finalize and persist one cycle to the trace file."""
cycle.timestamp_end = datetime.now(timezone.utc).isoformat()
# Compute duration
try:
t0 = datetime.fromisoformat(cycle.timestamp_start)
t1 = datetime.fromisoformat(cycle.timestamp_end)
cycle.duration_ms = int((t1 - t0).total_seconds() * 1000)
except (ValueError, TypeError):
cycle.duration_ms = 0
# Count successes/failures
cycle.actions_succeeded = sum(
1 for a in cycle.actions_executed if a.get("success", False)
)
cycle.actions_failed = sum(
1 for a in cycle.actions_executed if not a.get("success", True)
)
self.cycles.append(cycle)
# Append to JSONL
with open(self.trace_file, "a") as f:
f.write(json.dumps(cycle.to_dict()) + "\n")
def finish_session(self) -> SessionManifest:
"""Finalize the session and write the manifest."""
self.finished_at = datetime.now(timezone.utc).isoformat()
total_actions = sum(len(c.actions_executed) for c in self.cycles)
total_succeeded = sum(c.actions_succeeded for c in self.cycles)
total_failed = sum(c.actions_failed for c in self.cycles)
manifest = SessionManifest(
trace_id=self.trace_id,
harness_session_id=self.harness_session_id,
hermes_session_id=self.hermes_session_id,
hermes_log_id=self.hermes_log_id,
started_at=self.started_at,
finished_at=self.finished_at,
total_cycles=len(self.cycles),
total_actions=total_actions,
total_succeeded=total_succeeded,
total_failed=total_failed,
trace_file=str(self.trace_file),
trace_dir=str(self.trace_dir),
replay_command=(
f"python -m nexus.bannerlord_harness --mock --replay {self.trace_file}"
),
eval_note=(
"To replay: load this trace, re-execute each cycle's actions_planned "
"against a fresh harness in mock mode, compare actions_executed outcomes. "
"Success metric: >=90% action parity between original and replay runs."
),
)
with open(self.manifest_file, "w") as f:
json.dump(manifest.to_dict(), f, indent=2)
return manifest
@classmethod
def load_trace(cls, trace_file: Path) -> list[dict]:
"""Load a trace JSONL file for replay or analysis."""
cycles = []
with open(trace_file) as f:
for line in f:
line = line.strip()
if line:
cycles.append(json.loads(line))
return cycles
@classmethod
def load_manifest(cls, manifest_file: Path) -> dict:
"""Load a session manifest."""
with open(manifest_file) as f:
return json.load(f)
@classmethod
def list_traces(cls, trace_dir: Optional[Path] = None) -> list[dict]:
"""List all available trace sessions."""
d = trace_dir or DEFAULT_TRACE_DIR
if not d.exists():
return []
traces = []
for mf in sorted(d.glob("manifest_*.json")):
try:
manifest = cls.load_manifest(mf)
traces.append(manifest)
except (json.JSONDecodeError, IOError):
continue
return traces

View File

@@ -1,97 +0,0 @@
# Bannerlord Session Trace — Replay & Eval Guide
## Storage Layout
All traces live under `~/.timmy/traces/bannerlord/`:
```
~/.timmy/traces/bannerlord/
trace_<trace_id>.jsonl # One line per ODA cycle (full state + actions)
manifest_<trace_id>.json # Session metadata, counts, replay command
```
## Trace Format (JSONL)
Each line is one ODA cycle:
```json
{
"cycle_index": 0,
"timestamp_start": "2026-04-10T20:15:00+00:00",
"timestamp_end": "2026-04-10T20:15:45+00:00",
"duration_ms": 45000,
"screenshot_path": "/tmp/bannerlord_capture_1744320900.png",
"window_found": true,
"screen_size": [1920, 1080],
"mouse_position": [960, 540],
"playtime_hours": 142.5,
"players_online": 8421,
"is_running": true,
"actions_planned": [{"type": "move_to", "x": 960, "y": 540}],
"actions_executed": [{"success": true, "action": "move_to", ...}],
"actions_succeeded": 1,
"actions_failed": 0,
"hermes_session_id": "f47ac10b",
"hermes_log_id": "",
"harness_session_id": "f47ac10b"
}
```
## Capturing a Trace
```bash
# Run harness with trace logging enabled
cd /path/to/the-nexus
python -m nexus.bannerlord_harness --mock --trace --iterations 3
```
The trace and manifest are written to `~/.timmy/traces/bannerlord/` on harness shutdown.
## Replay Protocol
1. Load a trace: `BannerlordTraceLogger.load_trace(trace_file)`
2. Create a fresh harness in mock mode
3. For each cycle in the trace:
- Re-execute the `actions_planned` list
- Compare actual `actions_executed` outcomes against the recorded ones
4. Score: `(matching_actions / total_actions) * 100`
### Eval Criteria
| Score | Grade | Meaning |
|---------|----------|--------------------------------------------|
| >= 90% | PASS | Replay matches original closely |
| 70-89% | PARTIAL | Some divergence, investigate differences |
| < 70% | FAIL | Significant drift, review action semantics |
## Replay Script (sketch)
```python
from nexus.bannerlord_trace import BannerlordTraceLogger
from nexus.bannerlord_harness import BannerlordHarness
# Load trace
cycles = BannerlordTraceLogger.load_trace(
Path.home() / ".timmy" / "traces" / "bannerlord" / "trace_bl_xxx.jsonl"
)
# Replay
harness = BannerlordHarness(enable_mock=True, enable_trace=False)
await harness.start()
for cycle in cycles:
for action in cycle["actions_planned"]:
result = await harness.execute_action(action)
# Compare result against cycle["actions_executed"]
await harness.stop()
```
## Hermes Session Mapping
The `hermes_session_id` and `hermes_log_id` fields link traces to Hermes session logs.
When a trace is captured during a live Hermes session, populate these fields so
the trace can be correlated with the broader agent conversation context.

View File

@@ -1,18 +0,0 @@
{
"trace_id": "bl_20260410_201500_a1b2c3",
"harness_session_id": "f47ac10b",
"hermes_session_id": "f47ac10b",
"hermes_log_id": "",
"game": "Mount & Blade II: Bannerlord",
"app_id": 261550,
"started_at": "2026-04-10T20:15:00+00:00",
"finished_at": "2026-04-10T20:17:30+00:00",
"total_cycles": 3,
"total_actions": 6,
"total_succeeded": 6,
"total_failed": 0,
"trace_file": "~/.timmy/traces/bannerlord/trace_bl_20260410_201500_a1b2c3.jsonl",
"trace_dir": "~/.timmy/traces/bannerlord",
"replay_command": "python -m nexus.bannerlord_harness --mock --replay ~/.timmy/traces/bannerlord/trace_bl_20260410_201500_a1b2c3.jsonl",
"eval_note": "To replay: load trace, re-execute each cycle's actions_planned against a fresh harness in mock mode, compare actions_executed outcomes. Success metric: >=90% action parity between original and replay runs."
}

View File

@@ -1,3 +0,0 @@
{"cycle_index": 0, "timestamp_start": "2026-04-10T20:15:00+00:00", "timestamp_end": "2026-04-10T20:15:45+00:00", "duration_ms": 45000, "screenshot_path": "/tmp/bannerlord_capture_1744320900.png", "window_found": true, "screen_size": [1920, 1080], "mouse_position": [960, 540], "playtime_hours": 142.5, "players_online": 8421, "is_running": true, "actions_planned": [{"type": "move_to", "x": 960, "y": 540}, {"type": "press_key", "key": "space"}], "decision_note": "Initial state capture. Move to screen center and press space to advance.", "actions_executed": [{"success": true, "action": "move_to", "params": {"type": "move_to", "x": 960, "y": 540}, "timestamp": "2026-04-10T20:15:30+00:00", "error": null}, {"success": true, "action": "press_key", "params": {"type": "press_key", "key": "space"}, "timestamp": "2026-04-10T20:15:45+00:00", "error": null}], "actions_succeeded": 2, "actions_failed": 0, "hermes_session_id": "f47ac10b", "hermes_log_id": "", "harness_session_id": "f47ac10b"}
{"cycle_index": 1, "timestamp_start": "2026-04-10T20:15:45+00:00", "timestamp_end": "2026-04-10T20:16:30+00:00", "duration_ms": 45000, "screenshot_path": "/tmp/bannerlord_capture_1744320945.png", "window_found": true, "screen_size": [1920, 1080], "mouse_position": [960, 540], "playtime_hours": 142.5, "players_online": 8421, "is_running": true, "actions_planned": [{"type": "press_key", "key": "p"}], "decision_note": "Open party screen to inspect troops.", "actions_executed": [{"success": true, "action": "press_key", "params": {"type": "press_key", "key": "p"}, "timestamp": "2026-04-10T20:16:00+00:00", "error": null}], "actions_succeeded": 1, "actions_failed": 0, "hermes_session_id": "f47ac10b", "hermes_log_id": "", "harness_session_id": "f47ac10b"}
{"cycle_index": 2, "timestamp_start": "2026-04-10T20:16:30+00:00", "timestamp_end": "2026-04-10T20:17:30+00:00", "duration_ms": 60000, "screenshot_path": "/tmp/bannerlord_capture_1744321020.png", "window_found": true, "screen_size": [1920, 1080], "mouse_position": [960, 540], "playtime_hours": 142.5, "players_online": 8421, "is_running": true, "actions_planned": [{"type": "press_key", "key": "escape"}, {"type": "move_to", "x": 500, "y": 300}, {"type": "click", "x": 500, "y": 300}], "decision_note": "Close party screen, click on campaign map settlement.", "actions_executed": [{"success": true, "action": "press_key", "params": {"type": "press_key", "key": "escape"}, "timestamp": "2026-04-10T20:16:45+00:00", "error": null}, {"success": true, "action": "move_to", "params": {"type": "move_to", "x": 500, "y": 300}, "timestamp": "2026-04-10T20:17:00+00:00", "error": null}, {"success": true, "action": "click", "params": {"type": "click", "x": 500, "y": 300}, "timestamp": "2026-04-10T20:17:30+00:00", "error": null}], "actions_succeeded": 3, "actions_failed": 0, "hermes_session_id": "f47ac10b", "hermes_log_id": "", "harness_session_id": "f47ac10b"}