tribunal-kit 9.2.5 → 9.2.7

package/.agents/skills/hf-cloud-serving-image-selection/SKILL.md

@@ -1,10 +1,43 @@
----
-name: hf-cloud-serving-image-selection
-description: 'Pick the right serving container for a SageMaker model deployment and find its current image URI. Use this skill whenever about to deploy a model to a SageMaker endpoint and an image URI needs to be chosen — including when the user says "deploy this LLM", "host this HuggingFace model", "serve this fine-tuned model", "deploy this embedding model", "host a reranker", "serve a sentence-transformers model", or when about to hardcode any container URI in deployment code. HuggingFace-curated Deep Learning Containers are ALWAYS preferred: HuggingFace vLLM (LLMs and generative rerankers), HuggingFace vLLM-Omni (multimodal), TEI (embeddings/cross-encoder rerankers), HF Inference Toolkit (other transformers). Generic images (AWS vLLM, DJL-LMI, SGLang) are used only when no HuggingFace image is compatible — never merely because they carry a newer version. Never hardcode a container URI from memory and never default to TGI. Prevents stale-image failures and wrong-region URIs.'
----
-
-# Serving Image Selection
-
+---
+name: hf-cloud-serving-image-selection
+description: "Use when Pick the right serving container for a SageMaker model deployment and find its current image URI. Use this skill whenever about to deploy a model to a SageMaker endpoint and an image URI needs to be chosen — including when the user says \"deploy this LLM\", \"host this HuggingFace model\", \"serve this fine-tuned model\", \"deploy this embedding model\", \"host a reranker\", \"serve a sentence-transformers model\", or when about to hardcode any container URI in deployment code. HuggingFace-curated Deep Learning Containers are ALWAYS preferred: HuggingFace vLLM (LLMs and generative rerankers), HuggingFace vLLM-Omni (multimodal), TEI (embeddings/cross-encoder rerankers), HF Inference Toolkit (other transformers). Generic images (AWS vLLM, DJL-LMI, SGLang) are used only when no HuggingFace image is compatible — never merely because they carry a newer version. Never hardcode a container URI from memory and never default to TGI. Prevents stale-image failures and wrong-region URIs."
+version: 6.0.0
+last-updated: 2026-09-29
+tools: Read, Grep, Glob, Bash, Edit, Write
+scripts-binding:
+  - .agent/scripts/lint_runner.js
+  - .agent/scripts/verify_all.js
+---
+
+# Serving Image Selection
+
+## Mandatory Pre-Flight Context Inspection
+Before reading, generating, or refactoring code in the `hf-cloud-serving-image-selection` domain, inspect these 5 critical parameters:
+1. **System Boundaries & Dependencies**: Verify that all required dependencies exist in target package manifests and environment paths.
+2. **Runtime Context & Platform Invariants**: Confirm target platform constraints (Node.js, Browser, Mobile OS, Edge runtime) before applying APIs.
+3. **Execution Guardrails**: Identify potential side-effects, state mutations, and unhandled asynchronous exceptions.
+4. **Validation & Type Contracts**: Validate input data schemas and strict type constraints across all module interfaces.
+5. **Observability & Proof of Execution**: Ensure execution produces tangible verification signals (terminal output, tests, metrics).
+
+
+## Activation Boundaries
+- **Activate when:** Use when Pick the right serving container for a SageMaker model deployment and find its current image URI. Use this skill whenever about to deploy a model to a SageMaker endpoint and an image URI needs to be chosen — including when the user says "deploy this LLM", "host this HuggingFace model", "serve this fine-tuned model", "deploy this embedding model", "host a reranker", "serve a sentence-transformers model", or when about to hardcode any container URI in deployment code. HuggingFace-curated Deep Learning Containers are ALWAYS preferred: HuggingFace vLLM (LLMs and generative rerankers), HuggingFace vLLM-Omni (multimodal), TEI (embeddings/cross-encoder rerankers), HF Inference Toolkit (other transformers). Generic images (AWS vLLM, DJL-LMI, SGLang) are used only when no HuggingFace image is compatible — never merely because they carry a newer version. Never hardcode a container URI from memory and never default to TGI. Prevents stale-image failures and wrong-region URIs.
+- **DO NOT activate when:** The task falls outside the `hf-cloud-serving-image-selection` domain or is managed by a different dedicated specialist agent.
+
+
+## 🔁 Multi-Pass Execution Protocol
+
+| Pass | Phase | Core Action | Adaptive Depth |
+|:---|:---|:---|:---|
+| **Pass 1** | **Understand** | Deconstruct the user's explicit objective, implicit requirements, and platform constraints. | Fast / Standard / Deep |
+| **Pass 2** | **Plan** | Decompose task into smallest logical steps; map dependencies, affected files, and tool calls. | Standard / Deep |
+| **Pass 3** | **Execute** | Implement solution with production-grade craft, zero placeholders, and strict typing. | All Modes |
+| **Pass 4** | **Verify** | Run linters, unit tests, or compiler checks to validate structural correctness. | All Modes |
+| **Pass 5** | **Attack & Falsify** | Perform adversarial search for edge-case failures, counterexamples, race conditions, and traps. | Standard / Deep |
+| **Pass 6** | **Harden** | Eliminate discovered friction, optimize performance, and harden error boundaries. | Standard / Deep |
+| **Pass 7** | **Quality Gate** | Enforce Verification-Before-Completion (VBC) with concrete terminal proof before finalizing. | All Modes |
+
+
 The serving container is the single thing most likely to break a SageMaker deployment that "looked correct on paper". Wrong container, stale tag, or the wrong AMI — all produce the same opaque `Failed to pass health check` error.
 
 ## Rule zero: HuggingFace images always win
@@ -217,4 +250,49 @@ Re-check when AWS publishes new `huggingface-pytorch-inference` GPU tags — rem
 
 Loading the model from HF Hub happens *inside the container after the endpoint starts* — expect **5–15+ minutes** before InService even for small models, longer for multi-GB ones. A slow first boot is not a failure; don't tear down or re-diagnose before the deploy script's 30-minute wait expires.
 
-For production or repeated deployments, pre-stage the weights in S3 and pass `--model-s3-uri` to `deploy.py` (the model then loads from `/opt/ml/model`) — faster, immune to Hub rate limits/outages, and no `HUGGING_FACE_HUB_TOKEN` needed at runtime.
+For production or repeated deployments, pre-stage the weights in S3 and pass `--model-s3-uri` to `deploy.py` (the model then loads from `/opt/ml/model`) — faster, immune to Hub rate limits/outages, and no `HUGGING_FACE_HUB_TOKEN` needed at runtime.
+
+## 🚨 Edge-Case & Failure Mode Matrix
+
+| Scenario | Risk | Production Mitigation |
+|:---|:---|:---|
+| **Empty or Null Inputs** | Unhandled exception or unexpected rendering collapse | Enforce fallback guards, optional chaining, and explicit empty state handlers |
+| **Network Timeout / Latency** | Hanging operations or duplicate side-effects | Implement bounded abort controllers, exponential backoff, and idempotency keys |
+| **Concurrency / Race Conditions** | Stale state overwrite or inconsistent data mutations | Use atomic transactions, mutex locking, or cancel-on-resubmit controls |
+| **Invalid Schema / Malformed Payload** | Downstream runtime errors or security injection | Validate boundary payloads with Zod/Pydantic schemas prior to execution |
+| **Resource / Memory Saturation** | OOM errors, frame drops, or memory leaks | Clean up listeners, cancel active timers, and enforce pagination/virtualization |
+
+
+## 🤖 LLM-Specific Traps Table
+
+| Anti-Pattern | What AI Commonly Does Wrong | What Is Actually Correct |
+|:---|:---|:---|
+| **Platform-Dependent Shell Commands** | Hardcoding $(cat ~/.cache/huggingface/token) or piping to jq | Use platform-agnostic token checks via env var HF_TOKEN and native JSON parsing |
+| **VRAM OOM Crash** | Loading 70B model weights on a consumer 16GB GPU without quantization | Calculate parameter bounds (fp16 = VRAM/2, Q4 = VRAM*2) and load with bitsandbytes/GGUF |
+| **Blocking Batch Generation** | Generating large batches synchronously without streaming or yield | Use TextIteratorStreamer or asynchronous generate workers to prevent request timeouts |
+
+
+## 🏛️ Tribunal Verification & Guardrails
+
+**Active Reviewers:** `ai-code-reviewer` · `python-pro` · `performance-optimizer`
+**Slash Command:** `/review` or `/tribunal-full`
+
+### 🔬 Evidence Standard (Tri-State Verification)
+Every finding, audit statement, or completion claim must classify its factual certainty:
+- **`[OBSERVED]`**: Directly confirmed in the codebase or verified via executed terminal command.
+- **`[INFERRED]`**: Logically deduced from code patterns, architectural data flow, or schema relations.
+- **`[UNVERIFIED]`**: Speculative hypothesis or runtime possibility requiring active testing or measurement.
+
+### ✅ Pre-Flight Self-Audit Checklist
+```
+✅ Are model architectures and weights verified to fit within target hardware VRAM budgets?
+✅ Are pipeline requests guarded with bounded timeouts and retry backoffs?
+✅ Are dataset loading scripts operating in streaming mode to prevent out-of-memory errors?
+✅ Are API tokens and cache paths handled portably without assuming UNIX shell environments?
+✅ Are tokenizers and generation parameters (max_new_tokens, temperature) strictly bounded?
+```
+
+### 🛑 Verification-Before-Completion (VBC) Protocol
+**CRITICAL:** You must follow a strict "evidence-based closeout" state machine.
+- ❌ **Forbidden:** Declaring a task complete because the output "looks correct."
+- ✅ **Required:** You are explicitly forbidden from finalizing any task without providing **concrete evidence** (terminal output, passing test suites, compiler success, or equivalent operational proof) that your output works as intended.

package/.agents/skills/hf-mem/SKILL.md

@@ -1,8 +1,43 @@
----
-name: hf-mem
-description: Hugging Face CLI to estimate the required memory to load Safetensors or GGUF model weights for inference from the Hugging Face Hub
----
-
+---
+name: hf-mem
+description: "Use when Hugging Face CLI to estimate the required memory to load Safetensors or GGUF model weights for inference from the Hugging Face Hub"
+version: 6.0.0
+last-updated: 2026-09-29
+tools: Read, Grep, Glob, Bash, Edit, Write
+scripts-binding:
+  - .agent/scripts/lint_runner.js
+  - .agent/scripts/verify_all.js
+---
+
+# Hf Mem Engineering
+
+## Mandatory Pre-Flight Context Inspection
+Before reading, generating, or refactoring code in the `hf-mem` domain, inspect these 5 critical parameters:
+1. **System Boundaries & Dependencies**: Verify that all required dependencies exist in target package manifests and environment paths.
+2. **Runtime Context & Platform Invariants**: Confirm target platform constraints (Node.js, Browser, Mobile OS, Edge runtime) before applying APIs.
+3. **Execution Guardrails**: Identify potential side-effects, state mutations, and unhandled asynchronous exceptions.
+4. **Validation & Type Contracts**: Validate input data schemas and strict type constraints across all module interfaces.
+5. **Observability & Proof of Execution**: Ensure execution produces tangible verification signals (terminal output, tests, metrics).
+
+
+## Activation Boundaries
+- **Activate when:** Use when Hugging Face CLI to estimate the required memory to load Safetensors or GGUF model weights for inference from the Hugging Face Hub
+- **DO NOT activate when:** The task falls outside the `hf-mem` domain or is managed by a different dedicated specialist agent.
+
+
+## 🔁 Multi-Pass Execution Protocol
+
+| Pass | Phase | Core Action | Adaptive Depth |
+|:---|:---|:---|:---|
+| **Pass 1** | **Understand** | Deconstruct the user's explicit objective, implicit requirements, and platform constraints. | Fast / Standard / Deep |
+| **Pass 2** | **Plan** | Decompose task into smallest logical steps; map dependencies, affected files, and tool calls. | Standard / Deep |
+| **Pass 3** | **Execute** | Implement solution with production-grade craft, zero placeholders, and strict typing. | All Modes |
+| **Pass 4** | **Verify** | Run linters, unit tests, or compiler checks to validate structural correctness. | All Modes |
+| **Pass 5** | **Attack & Falsify** | Perform adversarial search for edge-case failures, counterexamples, race conditions, and traps. | Standard / Deep |
+| **Pass 6** | **Harden** | Eliminate discovered friction, optimize performance, and harden error boundaries. | Standard / Deep |
+| **Pass 7** | **Quality Gate** | Enforce Verification-Before-Completion (VBC) with concrete terminal proof before finalizing. | All Modes |
+
+
 `hf_mem` estimates the required memory for inference, including model weights and an optional KV cache, for Safetensors and GGUF for models on the Hugging Face Hub using HTTP Range requests i.e., without downloading or loading any weights locally.
 
 ## When to use?
@@ -76,4 +111,49 @@ And, for LLMs or VLMs with GGUF weights:
 
 ```bash
 uvx hf-mem --model-id unsloth/Qwen3.5-397B-A17B-GGUF --gguf-file Q4_K_M --experimental --json-output
-```
+```
+
+## 🚨 Edge-Case & Failure Mode Matrix
+
+| Scenario | Risk | Production Mitigation |
+|:---|:---|:---|
+| **Empty or Null Inputs** | Unhandled exception or unexpected rendering collapse | Enforce fallback guards, optional chaining, and explicit empty state handlers |
+| **Network Timeout / Latency** | Hanging operations or duplicate side-effects | Implement bounded abort controllers, exponential backoff, and idempotency keys |
+| **Concurrency / Race Conditions** | Stale state overwrite or inconsistent data mutations | Use atomic transactions, mutex locking, or cancel-on-resubmit controls |
+| **Invalid Schema / Malformed Payload** | Downstream runtime errors or security injection | Validate boundary payloads with Zod/Pydantic schemas prior to execution |
+| **Resource / Memory Saturation** | OOM errors, frame drops, or memory leaks | Clean up listeners, cancel active timers, and enforce pagination/virtualization |
+
+
+## 🤖 LLM-Specific Traps Table
+
+| Anti-Pattern | What AI Commonly Does Wrong | What Is Actually Correct |
+|:---|:---|:---|
+| **Platform-Dependent Shell Commands** | Hardcoding $(cat ~/.cache/huggingface/token) or piping to jq | Use platform-agnostic token checks via env var HF_TOKEN and native JSON parsing |
+| **VRAM OOM Crash** | Loading 70B model weights on a consumer 16GB GPU without quantization | Calculate parameter bounds (fp16 = VRAM/2, Q4 = VRAM*2) and load with bitsandbytes/GGUF |
+| **Blocking Batch Generation** | Generating large batches synchronously without streaming or yield | Use TextIteratorStreamer or asynchronous generate workers to prevent request timeouts |
+
+
+## 🏛️ Tribunal Verification & Guardrails
+
+**Active Reviewers:** `ai-code-reviewer` · `python-pro` · `performance-optimizer`
+**Slash Command:** `/review` or `/tribunal-full`
+
+### 🔬 Evidence Standard (Tri-State Verification)
+Every finding, audit statement, or completion claim must classify its factual certainty:
+- **`[OBSERVED]`**: Directly confirmed in the codebase or verified via executed terminal command.
+- **`[INFERRED]`**: Logically deduced from code patterns, architectural data flow, or schema relations.
+- **`[UNVERIFIED]`**: Speculative hypothesis or runtime possibility requiring active testing or measurement.
+
+### ✅ Pre-Flight Self-Audit Checklist
+```
+✅ Are model architectures and weights verified to fit within target hardware VRAM budgets?
+✅ Are pipeline requests guarded with bounded timeouts and retry backoffs?
+✅ Are dataset loading scripts operating in streaming mode to prevent out-of-memory errors?
+✅ Are API tokens and cache paths handled portably without assuming UNIX shell environments?
+✅ Are tokenizers and generation parameters (max_new_tokens, temperature) strictly bounded?
+```
+
+### 🛑 Verification-Before-Completion (VBC) Protocol
+**CRITICAL:** You must follow a strict "evidence-based closeout" state machine.
+- ❌ **Forbidden:** Declaring a task complete because the output "looks correct."
+- ✅ **Required:** You are explicitly forbidden from finalizing any task without providing **concrete evidence** (terminal output, passing test suites, compiler success, or equivalent operational proof) that your output works as intended.

package/.agents/skills/huggingface-best/SKILL.md

@@ -1,18 +1,43 @@
----
-name: huggingface-best
-description: >-
-  Use when the user asks about finding the best, top, or recommended model for a task,
-  wants to know what AI model to use, or wants to compare models by benchmark scores.
-  Triggers on: "best model for X", "what model should I use for", "top models for [task]",
-  "which model runs on my laptop/machine/device", "recommend a model for", "what LLM should
-  I use for", "compare models for", "what's state of the art for", or any question about
-  choosing an AI model for a specific use case. Always use this skill when the user wants
-  model recommendations or comparisons, even if they don't explicitly mention HuggingFace
-  or benchmarks.
----
-
-# HuggingFace Best Model Finder
-
+---
+name: huggingface-best
+description: "Use when loading, evaluating, fine-tuning, and deploying huggingface best machine learning models, datasets, and inference pipelines."
+version: 6.0.0
+last-updated: 2026-09-29
+tools: Read, Grep, Glob, Bash, Edit, Write
+scripts-binding:
+  - .agent/scripts/lint_runner.js
+  - .agent/scripts/verify_all.js
+---
+
+# HuggingFace Best Model Finder
+
+## Mandatory Pre-Flight Context Inspection
+Before reading, generating, or refactoring code in the `huggingface-best` domain, inspect these 5 critical parameters:
+1. **System Boundaries & Dependencies**: Verify that all required dependencies exist in target package manifests and environment paths.
+2. **Runtime Context & Platform Invariants**: Confirm target platform constraints (Node.js, Browser, Mobile OS, Edge runtime) before applying APIs.
+3. **Execution Guardrails**: Identify potential side-effects, state mutations, and unhandled asynchronous exceptions.
+4. **Validation & Type Contracts**: Validate input data schemas and strict type constraints across all module interfaces.
+5. **Observability & Proof of Execution**: Ensure execution produces tangible verification signals (terminal output, tests, metrics).
+
+
+## Activation Boundaries
+- **Activate when:** Use when loading, evaluating, fine-tuning, and deploying huggingface best machine learning models, datasets, and inference pipelines.
+- **DO NOT activate when:** The task falls outside the `huggingface-best` domain or is managed by a different dedicated specialist agent.
+
+
+## 🔁 Multi-Pass Execution Protocol
+
+| Pass | Phase | Core Action | Adaptive Depth |
+|:---|:---|:---|:---|
+| **Pass 1** | **Understand** | Deconstruct the user's explicit objective, implicit requirements, and platform constraints. | Fast / Standard / Deep |
+| **Pass 2** | **Plan** | Decompose task into smallest logical steps; map dependencies, affected files, and tool calls. | Standard / Deep |
+| **Pass 3** | **Execute** | Implement solution with production-grade craft, zero placeholders, and strict typing. | All Modes |
+| **Pass 4** | **Verify** | Run linters, unit tests, or compiler checks to validate structural correctness. | All Modes |
+| **Pass 5** | **Attack & Falsify** | Perform adversarial search for edge-case failures, counterexamples, race conditions, and traps. | Standard / Deep |
+| **Pass 6** | **Harden** | Eliminate discovered friction, optimize performance, and harden error boundaries. | Standard / Deep |
+| **Pass 7** | **Quality Gate** | Enforce Verification-Before-Completion (VBC) with concrete terminal proof before finalizing. | All Modes |
+
+
 Finds the best models for a task by querying official HF benchmark leaderboards, enriching
 results with model size data, filtering for what fits on the user's device, and returning a
 comparison table with benchmark scores.
@@ -131,4 +156,49 @@ If they say yes, ask whether they'd prefer to:
 - **Leaderboard not found**: skip, note "leaderboard unavailable" in output
 - **Model missing from hub_repo_details**: fall back to parsing size from model name
 - **No benchmarks found for task**: use the curated fallback table above, or try `hub_repo_search` with `filters=["<task>"]` sorted by `trendingScore`
-- **All leaderboards fail**: fall back to `hub_repo_search` for popular models tagged with the task, note that results are by popularity rather than benchmark score
+- **All leaderboards fail**: fall back to `hub_repo_search` for popular models tagged with the task, note that results are by popularity rather than benchmark score
+
+## 🚨 Edge-Case & Failure Mode Matrix
+
+| Scenario | Risk | Production Mitigation |
+|:---|:---|:---|
+| **Empty or Null Inputs** | Unhandled exception or unexpected rendering collapse | Enforce fallback guards, optional chaining, and explicit empty state handlers |
+| **Network Timeout / Latency** | Hanging operations or duplicate side-effects | Implement bounded abort controllers, exponential backoff, and idempotency keys |
+| **Concurrency / Race Conditions** | Stale state overwrite or inconsistent data mutations | Use atomic transactions, mutex locking, or cancel-on-resubmit controls |
+| **Invalid Schema / Malformed Payload** | Downstream runtime errors or security injection | Validate boundary payloads with Zod/Pydantic schemas prior to execution |
+| **Resource / Memory Saturation** | OOM errors, frame drops, or memory leaks | Clean up listeners, cancel active timers, and enforce pagination/virtualization |
+
+
+## 🤖 LLM-Specific Traps Table
+
+| Anti-Pattern | What AI Commonly Does Wrong | What Is Actually Correct |
+|:---|:---|:---|
+| **Platform-Dependent Shell Commands** | Hardcoding $(cat ~/.cache/huggingface/token) or piping to jq | Use platform-agnostic token checks via env var HF_TOKEN and native JSON parsing |
+| **VRAM OOM Crash** | Loading 70B model weights on a consumer 16GB GPU without quantization | Calculate parameter bounds (fp16 = VRAM/2, Q4 = VRAM*2) and load with bitsandbytes/GGUF |
+| **Blocking Batch Generation** | Generating large batches synchronously without streaming or yield | Use TextIteratorStreamer or asynchronous generate workers to prevent request timeouts |
+
+
+## 🏛️ Tribunal Verification & Guardrails
+
+**Active Reviewers:** `ai-code-reviewer` · `python-pro` · `performance-optimizer`
+**Slash Command:** `/review` or `/tribunal-full`
+
+### 🔬 Evidence Standard (Tri-State Verification)
+Every finding, audit statement, or completion claim must classify its factual certainty:
+- **`[OBSERVED]`**: Directly confirmed in the codebase or verified via executed terminal command.
+- **`[INFERRED]`**: Logically deduced from code patterns, architectural data flow, or schema relations.
+- **`[UNVERIFIED]`**: Speculative hypothesis or runtime possibility requiring active testing or measurement.
+
+### ✅ Pre-Flight Self-Audit Checklist
+```
+✅ Are model architectures and weights verified to fit within target hardware VRAM budgets?
+✅ Are pipeline requests guarded with bounded timeouts and retry backoffs?
+✅ Are dataset loading scripts operating in streaming mode to prevent out-of-memory errors?
+✅ Are API tokens and cache paths handled portably without assuming UNIX shell environments?
+✅ Are tokenizers and generation parameters (max_new_tokens, temperature) strictly bounded?
+```
+
+### 🛑 Verification-Before-Completion (VBC) Protocol
+**CRITICAL:** You must follow a strict "evidence-based closeout" state machine.
+- ❌ **Forbidden:** Declaring a task complete because the output "looks correct."
+- ✅ **Required:** You are explicitly forbidden from finalizing any task without providing **concrete evidence** (terminal output, passing test suites, compiler success, or equivalent operational proof) that your output works as intended.

package/.agents/skills/huggingface-community-evals/SKILL.md

@@ -1,10 +1,43 @@
----
-name: huggingface-community-evals
-description: Run evaluations for Hugging Face Hub models using inspect-ai and lighteval on local hardware. Use for backend selection, local GPU evals, and choosing between vLLM / Transformers / accelerate. Not for HF Jobs orchestration, model-card PRs, .eval_results publication, or community-evals automation.
----
-
-# Overview
-
+---
+name: huggingface-community-evals
+description: "Use when Run evaluations for Hugging Face Hub models using inspect-ai and lighteval on local hardware. Use for backend selection, local GPU evals, and choosing between vLLM / Transformers / accelerate. Not for HF Jobs orchestration, model-card PRs, .eval_results publication, or community-evals automation."
+version: 6.0.0
+last-updated: 2026-09-29
+tools: Read, Grep, Glob, Bash, Edit, Write
+scripts-binding:
+  - .agent/scripts/lint_runner.js
+  - .agent/scripts/verify_all.js
+---
+
+# Overview
+
+## Mandatory Pre-Flight Context Inspection
+Before reading, generating, or refactoring code in the `huggingface-community-evals` domain, inspect these 5 critical parameters:
+1. **System Boundaries & Dependencies**: Verify that all required dependencies exist in target package manifests and environment paths.
+2. **Runtime Context & Platform Invariants**: Confirm target platform constraints (Node.js, Browser, Mobile OS, Edge runtime) before applying APIs.
+3. **Execution Guardrails**: Identify potential side-effects, state mutations, and unhandled asynchronous exceptions.
+4. **Validation & Type Contracts**: Validate input data schemas and strict type constraints across all module interfaces.
+5. **Observability & Proof of Execution**: Ensure execution produces tangible verification signals (terminal output, tests, metrics).
+
+
+## Activation Boundaries
+- **Activate when:** Use when Run evaluations for Hugging Face Hub models using inspect-ai and lighteval on local hardware. Use for backend selection, local GPU evals, and choosing between vLLM / Transformers / accelerate. Not for HF Jobs orchestration, model-card PRs, .eval_results publication, or community-evals automation.
+- **DO NOT activate when:** The task falls outside the `huggingface-community-evals` domain or is managed by a different dedicated specialist agent.
+
+
+## 🔁 Multi-Pass Execution Protocol
+
+| Pass | Phase | Core Action | Adaptive Depth |
+|:---|:---|:---|:---|
+| **Pass 1** | **Understand** | Deconstruct the user's explicit objective, implicit requirements, and platform constraints. | Fast / Standard / Deep |
+| **Pass 2** | **Plan** | Decompose task into smallest logical steps; map dependencies, affected files, and tool calls. | Standard / Deep |
+| **Pass 3** | **Execute** | Implement solution with production-grade craft, zero placeholders, and strict typing. | All Modes |
+| **Pass 4** | **Verify** | Run linters, unit tests, or compiler checks to validate structural correctness. | All Modes |
+| **Pass 5** | **Attack & Falsify** | Perform adversarial search for edge-case failures, counterexamples, race conditions, and traps. | Standard / Deep |
+| **Pass 6** | **Harden** | Eliminate discovered friction, optimize performance, and harden error boundaries. | Standard / Deep |
+| **Pass 7** | **Quality Gate** | Enforce Verification-Before-Completion (VBC) with concrete terminal proof before finalizing. | All Modes |
+
+
 This skill is for **running evaluations against models on the Hugging Face Hub on local hardware**.
 
 It covers:
@@ -204,4 +237,49 @@ See:
 - `examples/USAGE_EXAMPLES.md` for local command patterns
 - `scripts/inspect_eval_uv.py`
 - `scripts/inspect_vllm_uv.py`
-- `scripts/lighteval_vllm_uv.py`
+- `scripts/lighteval_vllm_uv.py`
+
+## 🚨 Edge-Case & Failure Mode Matrix
+
+| Scenario | Risk | Production Mitigation |
+|:---|:---|:---|
+| **Empty or Null Inputs** | Unhandled exception or unexpected rendering collapse | Enforce fallback guards, optional chaining, and explicit empty state handlers |
+| **Network Timeout / Latency** | Hanging operations or duplicate side-effects | Implement bounded abort controllers, exponential backoff, and idempotency keys |
+| **Concurrency / Race Conditions** | Stale state overwrite or inconsistent data mutations | Use atomic transactions, mutex locking, or cancel-on-resubmit controls |
+| **Invalid Schema / Malformed Payload** | Downstream runtime errors or security injection | Validate boundary payloads with Zod/Pydantic schemas prior to execution |
+| **Resource / Memory Saturation** | OOM errors, frame drops, or memory leaks | Clean up listeners, cancel active timers, and enforce pagination/virtualization |
+
+
+## 🤖 LLM-Specific Traps Table
+
+| Anti-Pattern | What AI Commonly Does Wrong | What Is Actually Correct |
+|:---|:---|:---|
+| **Platform-Dependent Shell Commands** | Hardcoding $(cat ~/.cache/huggingface/token) or piping to jq | Use platform-agnostic token checks via env var HF_TOKEN and native JSON parsing |
+| **VRAM OOM Crash** | Loading 70B model weights on a consumer 16GB GPU without quantization | Calculate parameter bounds (fp16 = VRAM/2, Q4 = VRAM*2) and load with bitsandbytes/GGUF |
+| **Blocking Batch Generation** | Generating large batches synchronously without streaming or yield | Use TextIteratorStreamer or asynchronous generate workers to prevent request timeouts |
+
+
+## 🏛️ Tribunal Verification & Guardrails
+
+**Active Reviewers:** `ai-code-reviewer` · `python-pro` · `performance-optimizer`
+**Slash Command:** `/review` or `/tribunal-full`
+
+### 🔬 Evidence Standard (Tri-State Verification)
+Every finding, audit statement, or completion claim must classify its factual certainty:
+- **`[OBSERVED]`**: Directly confirmed in the codebase or verified via executed terminal command.
+- **`[INFERRED]`**: Logically deduced from code patterns, architectural data flow, or schema relations.
+- **`[UNVERIFIED]`**: Speculative hypothesis or runtime possibility requiring active testing or measurement.
+
+### ✅ Pre-Flight Self-Audit Checklist
+```
+✅ Are model architectures and weights verified to fit within target hardware VRAM budgets?
+✅ Are pipeline requests guarded with bounded timeouts and retry backoffs?
+✅ Are dataset loading scripts operating in streaming mode to prevent out-of-memory errors?
+✅ Are API tokens and cache paths handled portably without assuming UNIX shell environments?
+✅ Are tokenizers and generation parameters (max_new_tokens, temperature) strictly bounded?
+```
+
+### 🛑 Verification-Before-Completion (VBC) Protocol
+**CRITICAL:** You must follow a strict "evidence-based closeout" state machine.
+- ❌ **Forbidden:** Declaring a task complete because the output "looks correct."
+- ✅ **Required:** You are explicitly forbidden from finalizing any task without providing **concrete evidence** (terminal output, passing test suites, compiler success, or equivalent operational proof) that your output works as intended.

package/.agents/skills/huggingface-datasets/SKILL.md

@@ -1,10 +1,43 @@
----
-name: huggingface-datasets
-description: Use this skill for Hugging Face Dataset Viewer API workflows that fetch subset/split metadata, paginate rows, search text, apply filters, download parquet URLs, and read size or statistics.
----
-
-# Hugging Face Dataset Viewer
-
+---
+name: huggingface-datasets
+description: "Use when Use this skill for Hugging Face Dataset Viewer API workflows that fetch subset/split metadata, paginate rows, search text, apply filters, download parquet URLs, and read size or statistics."
+version: 6.0.0
+last-updated: 2026-09-29
+tools: Read, Grep, Glob, Bash, Edit, Write
+scripts-binding:
+  - .agent/scripts/lint_runner.js
+  - .agent/scripts/verify_all.js
+---
+
+# Hugging Face Dataset Viewer
+
+## Mandatory Pre-Flight Context Inspection
+Before reading, generating, or refactoring code in the `huggingface-datasets` domain, inspect these 5 critical parameters:
+1. **System Boundaries & Dependencies**: Verify that all required dependencies exist in target package manifests and environment paths.
+2. **Runtime Context & Platform Invariants**: Confirm target platform constraints (Node.js, Browser, Mobile OS, Edge runtime) before applying APIs.
+3. **Execution Guardrails**: Identify potential side-effects, state mutations, and unhandled asynchronous exceptions.
+4. **Validation & Type Contracts**: Validate input data schemas and strict type constraints across all module interfaces.
+5. **Observability & Proof of Execution**: Ensure execution produces tangible verification signals (terminal output, tests, metrics).
+
+
+## Activation Boundaries
+- **Activate when:** Use when Use this skill for Hugging Face Dataset Viewer API workflows that fetch subset/split metadata, paginate rows, search text, apply filters, download parquet URLs, and read size or statistics.
+- **DO NOT activate when:** The task falls outside the `huggingface-datasets` domain or is managed by a different dedicated specialist agent.
+
+
+## 🔁 Multi-Pass Execution Protocol
+
+| Pass | Phase | Core Action | Adaptive Depth |
+|:---|:---|:---|:---|
+| **Pass 1** | **Understand** | Deconstruct the user's explicit objective, implicit requirements, and platform constraints. | Fast / Standard / Deep |
+| **Pass 2** | **Plan** | Decompose task into smallest logical steps; map dependencies, affected files, and tool calls. | Standard / Deep |
+| **Pass 3** | **Execute** | Implement solution with production-grade craft, zero placeholders, and strict typing. | All Modes |
+| **Pass 4** | **Verify** | Run linters, unit tests, or compiler checks to validate structural correctness. | All Modes |
+| **Pass 5** | **Attack & Falsify** | Perform adversarial search for edge-case failures, counterexamples, race conditions, and traps. | Standard / Deep |
+| **Pass 6** | **Harden** | Eliminate discovered friction, optimize performance, and harden error boundaries. | Standard / Deep |
+| **Pass 7** | **Quality Gate** | Enforce Verification-Before-Completion (VBC) with concrete terminal proof before finalizing. | All Modes |
+
+
 Use this skill to execute read-only Dataset Viewer API calls for dataset exploration and extraction.
 
 ## Core workflow
@@ -104,4 +137,49 @@ Default to private dataset repos because traces can contain prompts, file paths,
 ```bash
 hf repos create <namespace>/<repo> --type dataset --private --exist-ok
 hf upload <namespace>/<repo> ~/.codex/sessions codex/<project-or-cwd> --type dataset
-```
+```
+
+## 🚨 Edge-Case & Failure Mode Matrix
+
+| Scenario | Risk | Production Mitigation |
+|:---|:---|:---|
+| **Empty or Null Inputs** | Unhandled exception or unexpected rendering collapse | Enforce fallback guards, optional chaining, and explicit empty state handlers |
+| **Network Timeout / Latency** | Hanging operations or duplicate side-effects | Implement bounded abort controllers, exponential backoff, and idempotency keys |
+| **Concurrency / Race Conditions** | Stale state overwrite or inconsistent data mutations | Use atomic transactions, mutex locking, or cancel-on-resubmit controls |
+| **Invalid Schema / Malformed Payload** | Downstream runtime errors or security injection | Validate boundary payloads with Zod/Pydantic schemas prior to execution |
+| **Resource / Memory Saturation** | OOM errors, frame drops, or memory leaks | Clean up listeners, cancel active timers, and enforce pagination/virtualization |
+
+
+## 🤖 LLM-Specific Traps Table
+
+| Anti-Pattern | What AI Commonly Does Wrong | What Is Actually Correct |
+|:---|:---|:---|
+| **Platform-Dependent Shell Commands** | Hardcoding $(cat ~/.cache/huggingface/token) or piping to jq | Use platform-agnostic token checks via env var HF_TOKEN and native JSON parsing |
+| **VRAM OOM Crash** | Loading 70B model weights on a consumer 16GB GPU without quantization | Calculate parameter bounds (fp16 = VRAM/2, Q4 = VRAM*2) and load with bitsandbytes/GGUF |
+| **Blocking Batch Generation** | Generating large batches synchronously without streaming or yield | Use TextIteratorStreamer or asynchronous generate workers to prevent request timeouts |
+
+
+## 🏛️ Tribunal Verification & Guardrails
+
+**Active Reviewers:** `ai-code-reviewer` · `python-pro` · `performance-optimizer`
+**Slash Command:** `/review` or `/tribunal-full`
+
+### 🔬 Evidence Standard (Tri-State Verification)
+Every finding, audit statement, or completion claim must classify its factual certainty:
+- **`[OBSERVED]`**: Directly confirmed in the codebase or verified via executed terminal command.
+- **`[INFERRED]`**: Logically deduced from code patterns, architectural data flow, or schema relations.
+- **`[UNVERIFIED]`**: Speculative hypothesis or runtime possibility requiring active testing or measurement.
+
+### ✅ Pre-Flight Self-Audit Checklist
+```
+✅ Are model architectures and weights verified to fit within target hardware VRAM budgets?
+✅ Are pipeline requests guarded with bounded timeouts and retry backoffs?
+✅ Are dataset loading scripts operating in streaming mode to prevent out-of-memory errors?
+✅ Are API tokens and cache paths handled portably without assuming UNIX shell environments?
+✅ Are tokenizers and generation parameters (max_new_tokens, temperature) strictly bounded?
+```
+
+### 🛑 Verification-Before-Completion (VBC) Protocol
+**CRITICAL:** You must follow a strict "evidence-based closeout" state machine.
+- ❌ **Forbidden:** Declaring a task complete because the output "looks correct."
+- ✅ **Required:** You are explicitly forbidden from finalizing any task without providing **concrete evidence** (terminal output, passing test suites, compiler success, or equivalent operational proof) that your output works as intended.

package/.agents/skills/huggingface-gradio/SKILL.md

@@ -1,10 +1,43 @@
----
-name: huggingface-gradio
-description: Build Gradio web UIs and demos in Python. Use when creating or editing Gradio apps, components, event listeners, layouts, or chatbots.
----
-
-# Gradio
-
+---
+name: huggingface-gradio
+description: "Use when Build Gradio web UIs and demos in Python. Use when creating or editing Gradio apps, components, event listeners, layouts, or chatbots."
+version: 6.0.0
+last-updated: 2026-09-29
+tools: Read, Grep, Glob, Bash, Edit, Write
+scripts-binding:
+  - .agent/scripts/lint_runner.js
+  - .agent/scripts/verify_all.js
+---
+
+# Gradio
+
+## Mandatory Pre-Flight Context Inspection
+Before reading, generating, or refactoring code in the `huggingface-gradio` domain, inspect these 5 critical parameters:
+1. **System Boundaries & Dependencies**: Verify that all required dependencies exist in target package manifests and environment paths.
+2. **Runtime Context & Platform Invariants**: Confirm target platform constraints (Node.js, Browser, Mobile OS, Edge runtime) before applying APIs.
+3. **Execution Guardrails**: Identify potential side-effects, state mutations, and unhandled asynchronous exceptions.
+4. **Validation & Type Contracts**: Validate input data schemas and strict type constraints across all module interfaces.
+5. **Observability & Proof of Execution**: Ensure execution produces tangible verification signals (terminal output, tests, metrics).
+
+
+## Activation Boundaries
+- **Activate when:** Use when Build Gradio web UIs and demos in Python. Use when creating or editing Gradio apps, components, event listeners, layouts, or chatbots.
+- **DO NOT activate when:** The task falls outside the `huggingface-gradio` domain or is managed by a different dedicated specialist agent.
+
+
+## 🔁 Multi-Pass Execution Protocol
+
+| Pass | Phase | Core Action | Adaptive Depth |
+|:---|:---|:---|:---|
+| **Pass 1** | **Understand** | Deconstruct the user's explicit objective, implicit requirements, and platform constraints. | Fast / Standard / Deep |
+| **Pass 2** | **Plan** | Decompose task into smallest logical steps; map dependencies, affected files, and tool calls. | Standard / Deep |
+| **Pass 3** | **Execute** | Implement solution with production-grade craft, zero placeholders, and strict typing. | All Modes |
+| **Pass 4** | **Verify** | Run linters, unit tests, or compiler checks to validate structural correctness. | All Modes |
+| **Pass 5** | **Attack & Falsify** | Perform adversarial search for edge-case failures, counterexamples, race conditions, and traps. | Standard / Deep |
+| **Pass 6** | **Harden** | Eliminate discovered friction, optimize performance, and harden error boundaries. | Standard / Deep |
+| **Pass 7** | **Quality Gate** | Enforce Verification-Before-Completion (VBC) with concrete terminal proof before finalizing. | All Modes |
+
+
 Gradio is a Python library for building interactive web UIs and ML demos. This skill covers the core API, patterns, and examples.
 
 ## Guides
@@ -295,4 +328,49 @@ Both commands accept `--token` for accessing private Spaces.
 
 ## Additional Reference
 
-- [End-to-End Examples](examples.md) — complete working apps
+- [End-to-End Examples](examples.md) — complete working apps
+
+## 🚨 Edge-Case & Failure Mode Matrix
+
+| Scenario | Risk | Production Mitigation |
+|:---|:---|:---|
+| **Empty or Null Inputs** | Unhandled exception or unexpected rendering collapse | Enforce fallback guards, optional chaining, and explicit empty state handlers |
+| **Network Timeout / Latency** | Hanging operations or duplicate side-effects | Implement bounded abort controllers, exponential backoff, and idempotency keys |
+| **Concurrency / Race Conditions** | Stale state overwrite or inconsistent data mutations | Use atomic transactions, mutex locking, or cancel-on-resubmit controls |
+| **Invalid Schema / Malformed Payload** | Downstream runtime errors or security injection | Validate boundary payloads with Zod/Pydantic schemas prior to execution |
+| **Resource / Memory Saturation** | OOM errors, frame drops, or memory leaks | Clean up listeners, cancel active timers, and enforce pagination/virtualization |
+
+
+## 🤖 LLM-Specific Traps Table
+
+| Anti-Pattern | What AI Commonly Does Wrong | What Is Actually Correct |
+|:---|:---|:---|
+| **Platform-Dependent Shell Commands** | Hardcoding $(cat ~/.cache/huggingface/token) or piping to jq | Use platform-agnostic token checks via env var HF_TOKEN and native JSON parsing |
+| **VRAM OOM Crash** | Loading 70B model weights on a consumer 16GB GPU without quantization | Calculate parameter bounds (fp16 = VRAM/2, Q4 = VRAM*2) and load with bitsandbytes/GGUF |
+| **Blocking Batch Generation** | Generating large batches synchronously without streaming or yield | Use TextIteratorStreamer or asynchronous generate workers to prevent request timeouts |
+
+
+## 🏛️ Tribunal Verification & Guardrails
+
+**Active Reviewers:** `ai-code-reviewer` · `python-pro` · `performance-optimizer`
+**Slash Command:** `/review` or `/tribunal-full`
+
+### 🔬 Evidence Standard (Tri-State Verification)
+Every finding, audit statement, or completion claim must classify its factual certainty:
+- **`[OBSERVED]`**: Directly confirmed in the codebase or verified via executed terminal command.
+- **`[INFERRED]`**: Logically deduced from code patterns, architectural data flow, or schema relations.
+- **`[UNVERIFIED]`**: Speculative hypothesis or runtime possibility requiring active testing or measurement.
+
+### ✅ Pre-Flight Self-Audit Checklist
+```
+✅ Are model architectures and weights verified to fit within target hardware VRAM budgets?
+✅ Are pipeline requests guarded with bounded timeouts and retry backoffs?
+✅ Are dataset loading scripts operating in streaming mode to prevent out-of-memory errors?
+✅ Are API tokens and cache paths handled portably without assuming UNIX shell environments?
+✅ Are tokenizers and generation parameters (max_new_tokens, temperature) strictly bounded?
+```
+
+### 🛑 Verification-Before-Completion (VBC) Protocol
+**CRITICAL:** You must follow a strict "evidence-based closeout" state machine.
+- ❌ **Forbidden:** Declaring a task complete because the output "looks correct."
+- ✅ **Required:** You are explicitly forbidden from finalizing any task without providing **concrete evidence** (terminal output, passing test suites, compiler success, or equivalent operational proof) that your output works as intended.