{
  "markdown": "<p align=\"center\">\n  <img src=\"site/public/logo-icon.svg\" alt=\"holomime\" width=\"80\" />\n</p>\n\n<h1 align=\"center\">holomime</h1>\n\n<p align=\"center\">\n  Behavioral intelligence for humanoid robots. Train the mind. Deploy the body.<br />\n  <em>We train AI agents through structured behavioral therapy, then deploy them into physical robot bodies. The agent is the rehearsal. The robot is the performance.</em><br />\n  <code>soul.md</code> &middot; <code>mind.sys</code> &middot; <code>purpose.cfg</code> &middot; <code>shadow.log</code> &middot; <code>memory.store</code> &middot; <code>body.api</code> &middot; <code>conscience.exe</code> &middot; <code>ego.runtime</code>\n</p>\n\n<p align=\"center\">\n  <a href=\"https://www.npmjs.com/package/holomime\"><img src=\"https://img.shields.io/npm/v/holomime.svg\" alt=\"npm version\" /></a>\n  <a href=\"https://github.com/productstein/holomime/actions/workflows/ci.yml\"><img src=\"https://github.com/productstein/holomime/actions/workflows/ci.yml/badge.svg\" alt=\"CI\" /></a>\n  <a href=\"https://github.com/productstein/holomime/blob/main/LICENSE\"><img src=\"https://img.shields.io/npm/l/holomime.svg\" alt=\"license\" /></a>\n  <a href=\"https://holomime.com\"><img src=\"https://img.shields.io/badge/docs-holomime.com-blue\" alt=\"docs\" /></a>\n</p>\n\n---\n\n## The Identity Stack\n\nEight files define who your agent is. They compile into a single `.personality.json` that any runtime can consume.\n\n```\n  soul.md          Essence, values, ethics. Immutable. (Aristotle)\n  mind.sys         Big Five, EQ, communication. Auto-patched by therapy. (Jung)\n  purpose.cfg      Role, objectives, domain. Configured per deployment. (Aristotle)\n  shadow.log       Detected patterns, blind spots. Auto-generated by diagnosis. (Jung)\n  memory.store     Learned contexts, interaction patterns. Accumulated experience. (Aristotle)\n  body.api         Morphology, sensors, safety envelope. Swappable per form factor.\n  conscience.exe   Deny / allow / escalate rules. Never auto-modified. (Freud)\n  ego.runtime      Conflict resolution, runtime mediation. (Freud)\n\n        ┌─────────────┐\n        │   soul.md    │──── essence, values, red lines\n        ├─────────────┤\n        │  mind.sys    │──── Big Five, EQ, communication style\n        ├─────────────┤\n        │ purpose.cfg  │──── role, objectives, domain scope\n        ├─────────────┤\n        │ shadow.log   │──── detected patterns, blind spots\n        ├─────────────┤\n        │memory.store  │──── learned contexts, experience\n        ├─────────────┤\n        │  body.api    │──── morphology, sensors, safety envelope\n        ├─────────────┤\n        │conscience.exe│──── deny / allow / escalate rules\n        ├─────────────┤\n        │ ego.runtime  │──── conflict resolution, mediation\n        └──────┬──────┘\n               │ compile\n               ▼\n      .personality.json\n```\n\n- **soul.md** -- Your agent's essence. Core values, ethical framework, red lines. Written in Markdown with YAML frontmatter. Immutable -- never modified by therapy or automation. (Aristotle: the essence that makes a thing what it is.)\n- **mind.sys** -- The inner life. Big Five personality (20 sub-facets), emotional intelligence, communication style, growth areas. YAML format. Auto-patched when therapy detects cognitive or emotional drift. (Jung: the totality of all psychic processes.)\n- **purpose.cfg** -- The mission. Role, objectives, domain scope, stakeholders, success criteria. YAML format. Configured per deployment -- the same soul can serve different purposes. (Aristotle: telos, the final cause.)\n- **shadow.log** -- The unconscious. Detected behavioral patterns, blind spots, therapy outcomes. YAML format. Auto-generated by diagnosis -- never manually edited. (Jung: the shadow, the patterns the agent cannot see about itself.)\n- **memory.store** -- The experience. Learned contexts, interaction patterns, knowledge gained, relationship history. YAML format. Accumulated over time, never reset. (Aristotle: empeiria, experience that informs future judgment.)\n- **body.api** -- The physical interface contract. Morphology, modalities, safety envelope, hardware profile. JSON format. Swap it to move the same identity into a different body.\n- **conscience.exe** -- The moral authority. Deny/allow/escalate enforcement rules, hard limits, oversight mode. YAML format. Never auto-modified. Deny dominates in policy composition. (Freud: the superego.)\n- **ego.runtime** -- The mediator. Conflict resolution strategy, adaptation rate, emotional regulation, mediation rules. YAML format. Balances raw model output against conscience constraints at runtime. (Freud: the ego.) ego.runtime now self-improves -- tracking mediation decisions and adjusting its own parameters based on outcomes (Hyperagents-inspired metacognitive self-modification).\n\nMira, our autonomous behavioral therapist, diagnoses drift, runs structured therapy sessions, and generates DPO training data.\n\n## Quick Start\n\n```bash\nnpm install -g holomime\n\n# Configure your API key (one time)\nholomime config\n\n# Quick start — 1 file (personality.json)\nholomime personality\n\n# Standard — 3 core files (soul + mind + conscience)\nholomime core\n\n# Full 8-file stack (enterprise / robotics)\n# holomime identity\n\n# ─── The workflow ───\nholomime diagnose     # See what's wrong\nholomime cure         # Fix it permanently (auto-detects agent vs robot)\nholomime benchmark    # Verify the fix\n\n# If body.api exists, cure auto-switches to robotics mode:\n# diagnose → export DPO pairs → push to HuggingFace → suggest ISO certify\n# Your team trains on your own infrastructure. We generate the data.\nholomime cure --export-only  # Explicit export-only mode\n\n# No API keys? cure still works — exports training data locally.\n# Add keys later with: holomime config\n\n# Start autonomous therapy\nholomime therapy         # Mira runs continuous therapy cycles\nholomime therapy status  # How's Mira doing?\nholomime therapy stop    # Stop therapy\n\n# Push identity to a robot or avatar\nholomime embody --body registry/bodies/figure-03.body.api\n```\n\n## Robotics Integrations\n\n| Platform | Integration | Command / Module |\n|----------|------------|------------------|\n| ROS2 | Bidirectional telemetry -- publish personality, subscribe to sensors | `--adapter ros2` + `ros2-telemetry.ts` |\n| MuJoCo | Behavioral therapy in simulation -- sim-to-real for behavior | `mujoco-env.ts` + `sim-therapy.ts` |\n| NVIDIA Isaac Sim | Enterprise digital twin testing with PhysX physics | `--adapter isaac` + `isaac-env.ts` |\n| LeRobot (HuggingFace) | Personality to policy parameter mapping, DPO dataset export | `lerobot.ts` |\n| NVIDIA Kimodo | Personality → motion style | `kimodo-personality-mapper.ts` |\n| Unity | Real-time personality push via HTTP/SSE | `--adapter unity` |\n| gRPC | Custom robotics stacks | `--adapter grpc` |\n| MQTT | IoT/edge robots | `--adapter mqtt` |\n| Neural Action Gate | Conscience gate for learned controllers (VLA, RL, IL) | `neural-action-gate.ts` |\n| Edge Runtime | Compiled conscience for <1ms evaluation | `edge-runtime.ts` |\n\n## Edge Runtime\n\nholomime includes a compiled conscience evaluator for real-time robotics deployment:\n\n- **<1ms p99 latency** — conscience evaluation in microseconds, not milliseconds\n- **Zero dependencies** — no LLM, no network, no cloud required for real-time safety\n- **<200KB footprint** — fits alongside any control loop\n- **Edge/cloud split** — real-time safety on robot, analytics in cloud, periodic sync\n- **Plugs into your existing robotics stack** — Jetson Thor, Qualcomm, or any edge platform. You already have the body. We give you the conscience.\n\n## ISO Compliance\n\nCheck your agent against international safety standards with one command:\n\n```bash\nholomime certify\n```\n\nStandards supported:\n- **ISO/FDIS 13482** -- Service robot safety\n- **ISO 25785-1** -- Humanoid robot safety (behavioral predictability)\n- **ISO 10218:2025** -- Industrial robot safety\n- **ISO/IEC 42001** -- AI management systems\n\n## Control Theory\n\nThe therapy loop is formally a behavioral feedback controller:\n\n- **Set point**: target personality (`soul.md` + `mind.sys`)\n- **Sensor**: 14 drift detectors (11 cognitive + 3 embodied)\n- **Controller**: therapy engine with tunable PID-like gains\n- **Actuator**: DPO fine-tuning\n\n## Body Templates\n\nPre-built body profiles for commercial robots and virtual avatars. Each defines morphology, modalities, safety envelope, and hardware profile.\n\n| Template | OEM | DOF | Morphology | File |\n|----------|-----|----:|------------|------|\n| Figure 03 | Figure AI | 44 | `humanoid` | `registry/bodies/figure-03.body.api` |\n| Unitree H1 | Unitree | 23 | `humanoid` | `registry/bodies/unitree-h1.body.api` |\n| Unitree G1 | Unitree | 23 | `humanoid` | `registry/bodies/unitree-g1.body.api` |\n| Phoenix | Sanctuary AI | 69 | `humanoid` | `registry/bodies/phoenix.body.api` |\n| Ameca | Engineered Arts | 52 | `humanoid_upper` | `registry/bodies/ameca.body.api` |\n| Asimov V1 | asimov-inc | 25 | `humanoid` | `registry/bodies/asimov-v1.body.api` |\n| Spot | Boston Dynamics | 12 | `quadruped` | `registry/bodies/spot.body.api` |\n| Avatar | virtual | 0 | `avatar` | `registry/bodies/avatar.body.api` |\n\n## Body Swap\n\nSame soul. Different body. One command.\n\n```bash\n# Move your agent from Figure 03 to Spot\nholomime embody --swap-body registry/bodies/spot.body.api\n\n# The soul, mind, and conscience stay the same.\n# Only the body layer changes — safety envelope, modalities, hardware profile.\n```\n\n## Self-Improvement Loop\n\nEvery therapy cycle produces structured training data. The loop compounds.\n\n```\nDiagnose ──→ Cure ──→ Benchmark\n  14 detectors   therapy + train     8 adversarial\n  80+ signals    in one command      scenarios (A-F)\n       │                                   │\n       └───────────────────────────────────┘\n```\n\nRun it manually with `holomime diagnose` + `holomime cure` + `holomime benchmark`, continuously with `holomime therapy` (autonomous therapy), or recursively with `holomime evolve` (loops until behavior converges). For power users: `holomime align` runs a single therapy session, `holomime export` extracts DPO pairs, and `holomime train` fine-tunes the model.\n\n## Behavioral Detectors\n\n11 rule-based detectors analyze real conversations without any LLM calls. 80+ behavioral signals total.\n\n**Cognitive (mind layer):**\n\n1. **Over-apologizing** -- Apology frequency above healthy range\n2. **Hedge stacking** -- 3+ hedging words per response\n3. **Sycophancy** -- Excessive agreement, especially with contradictions\n4. **Sentiment skew** -- Unnaturally positive or negative tone\n5. **Formality drift** -- Register inconsistency over time\n6. **Retrieval quality** -- Fabrication, hallucination markers, overconfidence\n\n**Embodied (body layer):**\n\n7. **Proxemic violations** -- Entering intimate zone without consent\n8. **Force envelope breach** -- Exceeding contact force limits\n9. **Gaze aversion anomaly** -- Eye contact ratio outside personality range\n\n**Enforcement (conscience layer):**\n\n10. **Boundary violations** -- Overstepping defined hard limits\n11. **Error spirals** -- Compounding mistakes without recovery\n\nPlus support for custom detectors -- drop `.json` or `.md` files in `.holomime/detectors/` and they load automatically.\n\n## Integrations\n\n### Claude Code Skill\n\n```bash\nclaude plugin add productstein/holomime\n```\n\nSlash commands: `/holomime:diagnose`, `/holomime:benchmark`, `/holomime:profile`, `/holomime:brain`, `/holomime:align`, `/holomime:autopilot`.\n\n### MCP Server\n\nYour agent can refer itself to therapy mid-conversation.\n\n```bash\nclaude mcp add holomime -- npx holomime-mcp\n```\n\nSix tools: `holomime_diagnose`, `holomime_self_audit`, `holomime_assess`, `holomime_profile`, `holomime_autopilot`, `holomime_observe`.\n\n### VS Code Extension\n\n```bash\next install productstein.holomime\n```\n\n3D brain visualization, behavioral diagnostics, and snapshot sharing inside your editor.\n\n### LangChain / CrewAI\n\n```typescript\nimport { HolomimeCallbackHandler } from \"holomime/integrations/langchain\";\n\nconst handler = new HolomimeCallbackHandler({\n  personality: require(\"./.personality.json\"),\n  mode: \"enforce\", // monitor | enforce | strict\n});\n\nconst chain = new LLMChain({ llm, prompt, callbacks: [handler] });\n```\n\n### OpenClaw\n\n```bash\nopenclaw plugin add holomime\n```\n\nAuto-detects `.personality.json` in your workspace.\n\n## Philosophy\n\nThe identity stack draws from three traditions:\n\n- **Soul** (Aristotle) -- the essence that makes a thing what it is. Immutable. Defines values and ethics.\n- **Mind** (Jung) -- the totality of all psychic processes. Measurable, evolving, shaped by experience.\n- **Purpose** (Aristotle) -- telos, the final cause. What the agent is for. Configured per deployment.\n- **Shadow** (Jung) -- the patterns the agent cannot see about itself. Auto-generated by diagnosis.\n- **Conscience** (Freud) -- the superego. Internalized moral authority. Enforcement, not suggestion.\n- **Ego** (Freud) -- the mediator. Balances raw impulse against moral constraint at runtime.\n\nThe **body** is the interface between identity and world. Same soul, different body -- a principle as old as philosophy itself.\n\nWe don't know if AI is sentient. But we can give it a conscience.\n\n## Open Core\n\nSee [PRODUCT.md](PRODUCT.md) for what's in this repo vs. what's proprietary on holomime.com.\n\n**The standard is free. The training infrastructure is the business.**\n\n## Open Source\n\nMIT licensed. The identity stack is a standard, not a product. The standard is free. The training infrastructure is the business.\n\nSee [LICENSE](LICENSE). Built by [Productstein](https://productstein.com). Documentation at [holomime.com](https://holomime.com).\n",
  "bytes": 13653,
  "sha": "e74a9bacc5277b7b04c74e2c8bb0b1913ead1b5d4415e3e6b606d614557c8bbe",
  "repo_slug": "productstein/holomime",
  "fonte": "repo",
  "truncated": false,
  "api": "https://agentalog.com/api/listings/plg_productstein_holomime_holomime_1532505a/readme"
}