Menรบ
Create a Model/System Card (MCARD) document through an interactive step-by-step flow. Guides the user through each section with specific questions and example responses.
Instalar con Codex o Claude Copia este prompt, pรฉgalo en Codex, Claude u otro asistente, y deja que revise la pรกgina de la skill y la instale por ti.
Basado en la clasificaciรณn ocupacional SOC
Execute an external audit of a Charter inside an auditor-side CLI (gemini-cli, claude-cli, copilot-cli, codex-cli). Reads the resolved audit prompt from the canonical location, audits with tool use, writes the report. Operator invokes one instance per CLI.
Generate the unified audit prompt for a Charter at the canonical filesystem location. The operator then opens N auditor-side CLIs (gemini-cli, claude-cli, copilot-cli, etc.) and invokes /straymark-audit-execute in each โ no copy/paste. Counterpart of /straymark-audit-review.
Consolidate N external auditor reports into a critical review document with verdicts, remediation plan, and auditor ratings. Then merge the external_audit YAML block into the Charter telemetry. Counterpart of /straymark-audit-prompt and /straymark-audit-execute.
Execute an external audit of a Charter inside an auditor-side CLI (gemini-cli, claude-cli, copilot-cli, codex-cli). Reads the resolved audit prompt from the canonical location, audits with tool use, writes the report. Operator invokes one instance per CLI.
Generate the unified audit prompt for a Charter at the canonical filesystem location. The operator then opens N auditor-side CLIs (gemini-cli, claude-cli, copilot-cli, etc.) and invokes /straymark-audit-execute in each โ no copy/paste. Counterpart of /straymark-audit-review.
Consolidate N external auditor reports into a critical review document with verdicts, remediation plan, and auditor ratings. Then merge the external_audit YAML block into the Charter telemetry. Counterpart of /straymark-audit-prompt and /straymark-audit-execute.
| name | straymark-mcard |
| description | Create a Model/System Card (MCARD) document through an interactive step-by-step flow. Guides the user through each section with specific questions and example responses. |
| allowed-tools | Read, Write, Glob, Bash(git status *, git log *, git diff *, date *, wc *) |
Create AI Model/System Card documentation through an interactive guided flow.
When invoked, follow these steps:
If the user specified a model name (e.g., /straymark-mcard GPT-4o), use it as the model name and proceed to step 2 asking only for the model type.
If no parameter is given, proceed to step 2 asking for both model name and type.
Ask the user:
โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ
โ StrayMark MCARD โ Model/System Card โ
โ โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฃ
โ โ
โ Step 1/7: Model Identity โ
โ โ
โ Please provide the following: โ
โ โ
โ 1. Model name: [e.g., "GPT-4o", "BERT-base-uncased"] โ
โ 2. Model type: โ
โ โข LLM โ Large Language Model โ
โ โข classifier โ Classification model โ
โ โข regressor โ Regression model โ
โ โข generator โ Generative model (image, audio, etc.) โ
โ โข recommender โ Recommendation system โ
โ โข other โ Specify โ
โ 3. Provider: [e.g., "OpenAI", "Google", "Hugging Face"] โ
โ 4. Version: [e.g., "2024-05-13", "v1.0", "gpt-4o-2024-05-13"] โ
โ 5. License: [e.g., "Proprietary", "Apache 2.0", "MIT"] โ
โ โ
โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ
Wait for user response before proceeding.
Ask the user:
โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ
โ Step 2/7: Intended Use โ
โ โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฃ
โ โ
โ 1. Primary intended uses: โ
โ Example: "Code generation", "Customer support chatbot", โ
โ "Sentiment analysis on product reviews" โ
โ โ
โ 2. Primary intended users: โ
โ Example: "Internal engineering team", "End users via API", โ
โ "Data science team" โ
โ โ
โ 3. Out-of-scope uses (what the model should NOT be used for): โ
โ Example: "Medical diagnosis", "Legal advice", โ
โ "Autonomous decision-making without human review" โ
โ โ
โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ
Wait for user response before proceeding.
Ask the user:
โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ
โ Step 3/7: Training Data โ
โ โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฃ
โ โ
โ 1. Dataset name: [e.g., "Common Crawl", "Internal corpus v3"] โ
โ 2. Source: [e.g., "Web scraping", "Licensed data provider"] โ
โ 3. Size: [e.g., "1.5T tokens", "500K samples", "200GB"] โ
โ 4. Collection methodology: โ
โ Example: "Web crawling with quality filters" โ
โ 5. Preprocessing steps: โ
โ Example: "Deduplication, PII removal, language filtering" โ
โ 6. Known limitations in the data: โ
โ Example: "English-centric, underrepresents African languages" โ
โ 7. PII assessment: โ
โ Example: "PII filtered using regex + NER; residual risk low" โ
โ 8. Data license: [e.g., "CC-BY-4.0", "Proprietary"] โ
โ โ
โ Type "unknown" or "N/A" for fields you cannot fill. โ
โ โ
โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ
Wait for user response before proceeding.
Ask the user:
โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ
โ Step 4/7: Performance & Evaluation โ
โ โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฃ
โ โ
โ Performance Metrics (provide at least one): โ
โ Format: Metric | Value | Test Dataset | Conditions โ
โ Example: "Accuracy | 92.3% | MMLU | 5-shot" โ
โ "F1 | 0.87 | SQuAD v2 | zero-shot" โ
โ "BLEU | 34.2 | WMT-22 en-de | beam search k=5" โ
โ โ
โ Disaggregated evaluation (optional): โ
โ Format: Subgroup | Metric | Value โ
โ Example: "English | Accuracy | 95.1%" โ
โ "Spanish | Accuracy | 88.4%" โ
โ โ
โ Type "unknown" if metrics are not available. โ
โ โ
โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ
Wait for user response before proceeding.
Ask the user:
โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ
โ Step 5/7: Bias, Security & Ethics โ
โ โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฃ
โ โ
โ Bias & Fairness (if evaluated): โ
โ Format: Group | Metric | Performance | Mitigation โ
โ Example: "Female | Toxicity | 0.02 | Content filter applied" โ
โ Type "not evaluated" if no bias analysis was performed. โ
โ โ
โ Security Concerns: โ
โ โข Known vulnerabilities: [None / describe] โ
โ โข Adversarial robustness: [Low / Medium / High] โ
โ โข Prompt injection risk: [Low / Medium / High / N/A] โ
โ โข Data poisoning risk: [Low / Medium / High] โ
โ โข Model extraction risk: [Low / Medium / High] โ
โ โ
โ Ethical Considerations: โ
โ โข Was sensitive data used in training? [Yes / No / Unknown] โ
โ โข Human subjects involved? [Yes / No / Unknown] โ
โ โข Dual-use potential? [Yes / No] โ
โ โข Societal impact notes: [Brief description] โ
โ โ
โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ
Wait for user response before proceeding.
Ask the user:
โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ
โ Step 6/7: Environmental Impact โ
โ โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฃ
โ โ
โ 1. Training energy (kWh): [e.g., "1,287 kWh", "Unknown"] โ
โ 2. CO2 equivalent (tons): [e.g., "0.58 tCO2", "Unknown"] โ
โ 3. Hardware used: [e.g., "8x NVIDIA A100 80GB"] โ
โ 4. Training duration: [e.g., "72 hours", "Unknown"] โ
โ 5. Inference cost: [e.g., "$0.005 per 1K tokens", "Unknown"] โ
โ 6. Region / Grid carbon intensity: [e.g., "US-East", "Unknown"] โ
โ โ
โ Type "unknown" for fields you cannot fill. โ
โ โ
โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ
Wait for user response before proceeding.
Ask the user:
โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ
โ Step 7/7: Limitations & Recommendations โ
โ โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฃ
โ โ
โ Known limitations: โ
โ Example: "Poor performance on low-resource languages", โ
โ "Context window limited to 128K tokens", โ
โ "Hallucination rate ~5% on factual queries" โ
โ โ
โ Known failure modes: โ
โ Example: "Produces incorrect math on multi-step problems", โ
โ "May refuse safe queries due to over-filtering" โ
โ โ
โ Recommendations for deployers: โ
โ Example: "Implement output filtering for PII", โ
โ "Use human-in-the-loop for high-stakes decisions", โ
โ "Monitor for performance drift monthly" โ
โ โ
โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ
Wait for user response before proceeding.
Based on the gathered information, determine:
EU AI Act risk level: Classify based on intended use:
unacceptable: social scoring, real-time biometric identificationhigh: employment decisions, credit scoring, law enforcement, educationlimited: chatbots, emotion recognition, deepfake generationminimal: spam filters, game AI, search optimizationnot_applicable: no EU deployment plannedNIST GenAI risks: Select applicable risks from:
privacy, bias, confabulation, cbrn, dangerous_content, environmental, human_ai_config, information_integrity, information_security, intellectual_property, obscene_content, value_chain
Overall risk level: low, medium, high, or critical
Read .straymark/config.yml to determine language:
language: en # or es
Use template path based on language:
en (default): .straymark/templates/TEMPLATE-MCARD.mdes: .straymark/templates/i18n/es/TEMPLATE-MCARD.mdDetermine the next sequence number:
# Get current date
date +%Y-%m-%d
# Find existing MCARD documents for today
ls .straymark/09-ai-models/MCARD-$(date +%Y-%m-%d)-*.md 2>/dev/null | wc -l
ID format: MCARD-YYYY-MM-DD-NNN
Display a summary and ask for confirmation:
โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ
โ StrayMark MCARD โ Summary โ
โ โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฃ
โ โ
โ Model: [model_name] ([model_type]) โ
โ Provider: [provider] โ Version: [version] โ
โ EU AI Act Risk: [risk classification] โ
โ NIST GenAI Risks: [applicable risks] โ
โ Overall Risk: [risk level] โ
โ โ
โ Proposed filename: โ
โ MCARD-YYYY-MM-DD-NNN-[model-name-slug].md โ
โ โ
โ Location: โ
โ .straymark/09-ai-models/ โ
โ โ
โ Review required: YES โ
โ โ
โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ
Confirm creation? [Y/n]:
Wait for user confirmation before proceeding.
YYYY-MM-DD -> Current dateNNN -> Sequence number (001, 002, etc.)[agent-name] -> your runtime's canonical agent identity (see AGENT-RULES.md ยง1 โ e.g. claude-code-v1.0, gemini-cli-v1.0, codex-cli-v1.0, cursor-v1.0; do not assume Claude)[Model Name] -> User-provided model namereview_required: true in the frontmatter.straymark/09-ai-models/MCARD-YYYY-MM-DD-NNN-[model-name-slug].mdAfter creation, display:
StrayMark MCARD created:
.straymark/09-ai-models/MCARD-YYYY-MM-DD-NNN-[description].md
Model: [model_name] ([model_type])
Provider: [provider]
Review required: yes
Risk level: [risk_level]
EU AI Act: [classification]
.straymark/09-ai-models/ directory: Create it before saving[To be determined][Information not publicly available] and note in limitations.straymark/config.yml: Default to English (en)Terminal compatibility: If the terminal does not support box-drawing characters (Unicode), use plain-text formatting with dashes and pipes instead (e.g.,
+--+instead ofโโโโ).