feat(compliance): add pii_masker skill using micro-f1-mask

[rosspeili]

Description

This PR introduces the compliance/pii_masker skill, a localized Privacy Firewall component utilizing the 270M parameter edge model arpacorp/micro-f1-mask over Ollama. It is engineered to transparently intercept, parse, and safely scrub Personally Identifiable Information (Names, Contextual Addresses, Crypto Wallets, Emails, etc.) from prompts before outward transmission to high-latency LLM providers.

The skill includes dynamic operational modes (mask, redact, remove) and operates statelessly, returning detected metadata schema constructs for programmatic session mapping. Included are global library updates, execution flow examples, robust schema testing, and a refined API configuration matrix.

Type of Change (Matches Issue Templates)

• 🚀 Skill Proposal: New Skill (Contains manifest.yaml, skill.py, and instructions.md)
• 🐛 Bug Report Fix: Non-breaking change which fixes an execution error or framework bug
• 📖 Doc Fix: Documentation Update
• 🧠 Framework Feature / RFC Updates: Core Framework Update (Changes to base_skill.py, loader.py, etc.)

Checklist (For the Submitting Agent / Developer)

• My code follows the Agent Code of Conduct.
• I have included a properly formatted manifest.yaml (if submitting a new skill).
• The skill logic operates purely in Python and does not rely on arbitrary LLM code generation.
• Requirements and env_vars are explicitly documented in the manifest.
• I have written unit tests proving deterministic execution and schema compliance.
• I have verified that SkillLoader successfully loads this module without missing dependency errors.

Constitution & Safety (If adding/modifying a Skill)

1. USER PRIVACY: Do not store or transmit sensitive PII externally; only use the specified local arpacorp/micro-f1-mask edge-model.
2. ACCURACY: In mask mode, perfectly preserve the structural tags (e.g., [PERSON_1]) for downstream replacement without mutating intent.
3. COMPLIANCE: Adhere strictly to the requested masking trajectory mode mapping.

Related Issues

Fixes #32