introspect

# Introspect A self-analysis pipeline that collects behavioral context, identifies patterns, and produces actionable insights as an HTML report. The pipeline has 4 phases, each building on the previous. ## Pipeline Overview | Phase | What | Mode | Output | |-------|------|------|--------| | 1. Collect | Gather data from 4 sources | 3 parallel inline Task invocations | Raw findings | | 2. Analyze | Synthesize into profile | AI + user dialogue | Strengths, costs, conflicts | | 3. Ground | Map to philosophical frameworks | Optional Analogia | Validated mapping | | 4. Report | Generate HTML report | Automated | `.html` file | If the user provides a specific question (e.g., "What are my curses?"), orient the analysis toward answering that question rather than producing a generic profile. --- ## Phase 1: Data Collection Launch 3 parallel ad-hoc inline Task(general-purpose) invocations. These are not pre-registered agent files — each is an inline prompt task launched for this pipeline only. Each collects from different sources and returns structured findings. All prompts are English per delegation rules; search keywords in quotes are exempt. ### Agent 1: Rules & Configuration ``` PURPOSE: Extract the user's explicitly stated preferences and constraints from their Claude Code configuration. Paths below written `{config_dir}/…` take `{config_dir}` = `CLAUDE_CONFIG_DIR` when set, else `~/.claude`. Resolve it ONCE per invocation with Bash `printf '%s\n' "${CLAUDE_CONFIG_DIR-$H...