total-recall

Run cross-model comparison of trigger phrase generation (haiku vs sonnet vs opus)

Remove trigger phrases for a file or pattern

Rebuild the master word-to-files reverse lookup index

Show all stored trigger phrases

Study a file or directory and generate semantic trigger phrases

Bootstrap triggers from docs, prompts, skills, and project documentation

Runs cross-model comparison — spawns study agents on haiku, sonnet, and opus to compare how model size affects trigger phrase generation. Analyzes per-model convergence and cross-model agreement.

Orchestrates trigger phrase generation using two-phase sampling — 5 study agents first, then 5 more only if convergence is weak. Finds patterns across samples to distill the most training-data-resonant phrase.

Reads a file and produces a single short phrase (1-6 words) capturing what the file is about. Designed to be spawned multiple times against the same file so that convergent associations across samples reveal training-data-resonant triggers.

Run cross-model comparison — scan files with haiku, sonnet, and opus study agents to see how model size affects trigger phrase generation.

Remove trigger phrases for a file or glob pattern from .claude/triggers.json

Build a reverse word-to-files index from all stored triggers. Model-aware — builds per-model lookup so agents find triggers matched to their own model.

Display all stored trigger phrases from .claude/triggers.json

Study files and generate semantic trigger phrases by spawning a researcher agent that samples model associations and finds convergent patterns. Supports model-specific triggers via --models flag.

Scan prompt files, CLAUDE.md docs, skill definitions, and other documentation to seed trigger phrases — bootstrapping total-recall from existing project knowledge. Supports --models flag for model-scoped triggers.