Enforces institutional coding and architectural rules for AI-generated code in PHP, Python, Magento 2 projects.
npx skills add https://github.com/infinri/Phaselock --skill phaselock
Please help me install this Agent Skill. Skill: phaselock Source: https://github.com/infinri/Phaselock/tree/main Command: npx skills add https://github.com/infinri/Phaselock --skill phaselock
AI code generation fails in predictable, recurring ways: monolithic output, skipped planning phases, self-auditing without verification, and tests as afterthoughts. Phaselock converts each failure mode into a hard rule with mechanical enforcement through hooks and verification scripts, ensuring AI agents generate, review, and verify code against institutional knowledge.
Use the phaselock skill to review your AI-generated code against the institutional coding bible and enforce all applicable architectural, security, and performance rules before committing.
High-intent search queries and answers about installing and using this skill.
AI code generation skips planning phases and produces monolithic output without verification. Enforcing coding standards converts each failure mode into a hard rule with mechanical enforcement hooks that block writes and verify output against architectural invariants before committing.
To stop AI code generation from skipping phases, implement a phased implementation protocol with sequential planning gates and test-first verification. This forces the AI to declare call-paths and domain invariants first, breaking monolithic output into dependency-ordered slices with per-slice validation.
Yes, you can enforce architectural rules for Magento 2 and PHP projects. The enforcement library includes 80 framework-specific rules across architecture, database, security, performance, and testing domains tailored for PHP, Python, and Magento 2 environments.
Sliced code generation is a technique that breaks implementation into dependency-ordered slices with per-slice validation. It prevents monolithic output by enforcing static analysis and handoff JSON validation between each slice, ensuring AI agents generate and verify code incrementally.
Yes, static analysis integration works with tiered workflow routing by classifying tasks into four tiers (Research, Patch, Standard, Complex). Each tier applies the appropriate level of ceremony, routing complex tasks to phased implementation gates with full static analysis verification.
AI-generated code fails architectural reviews because it self-audits without verification and treats tests as afterthoughts. Mechanical enforcement hooks prevent this by blocking writes that violate the 80 enforceable rules and requiring test-first verification against institutional coding knowledge.
JohnWebb4Define and persist Cursor rules in .cursor/rules/ using YAML frontmatter.
jordanhubbardAutomate pre-commit code review with static analysis and security scans.
gkganesh12Audit source code against documentation and test suites for architectural drift.
Tashima-TarshCreate custom regex-based hookify rules for Claude Code events.
