Category: Uncategorized

Context Browser-agent commands on Facebook timed out after 20 seconds. Root cause: the TM script had a hardcoded 20s per-command timeout, while FB pages take 60-120s per command due to heavy React rendering. Root Causes Found & Fixed 20s TM script timeout: Increased to 60s default. Server now passes CLI timeout to TM script (was […]

# Session Closeout: Live Dashboard at pezant.ca/dashboard (2026-04-17) ## Summary Built and deployed a live infrastructure dashboard at `pezant.ca/dashboard`, wiring up the existing frontend skeleton to real data sources and fixing the OAuth return-to-origin flow. ## What Was Done ### 1. Apache Proxy Configuration – Added `RewriteRule` for `/dashboard` → port 3003 in Apache config […]

Session Summary 3-hour session reactivating job search. Created a landscape survey of 35+ elite-caliber roles, established a quality bar, prepped OpenAI application materials, and identified Thinking Machines Lab as a new target. Key Outcomes Caliber bar established: 5-criteria filter (company prestige, seniority, comp trajectory, AI centrality, direct posting link) codified across pipeline config, rules, and […]

Context Full audit of the job pipeline system. The filter was too aggressive, excluding real AI roles at Sierra, Harvey, Anyscale, and Bay Area locations like Redwood City and Foster City. 254 roles were stuck at outreach_drafted with no nudge mechanism. No Discord or Sheets integration. What Changed Filter Fixes (highest impact) Location: Expanded allowlist […]

# Anthropic Recruiter Reply + Browser Agent Public Release **Date:** 2026-04-10 **Duration:** ~45 minutes **Repos touched:** assortedLLMTasks, claude-browser-agent (new) ## Context & Motivation Laura Small (Anthropic recruiter) reached out about the PM Consumer role referral. The HM had already reviewed Nick’s profile and leaned toward candidates with more founder experience, so the reply needed to […]

Context Browser-agent TM userscript was causing Edge to hang during extended sessions. Root cause: aggressive DOM polling (two uncoordinated loops at 2s and 3s), accumulated timeout timers, and server-side memory leaks from orphaned state. Fixes (v1.8.0) Merged polling loops: Two separate timers to single 3s tick(). Cuts DOM polling ~40%. Cached getPageState(): 2s TTL cache […]

# Auto Shorts: Analytics, Learning Agent & Experimentation Framework **Date:** 2026-04-09 **Duration:** ~4 hours **Repos touched:** auto-shorts, shorts-pipeline ## Context & Motivation The Auto Shorts system generates YouTube Shorts from long-form cooking videos for the Chef Agathe channels (EN + FR) but had zero visibility into performance. The user wanted to: 1. Measure how different […]

Context & Motivation Before fully transitioning Discord request handling from the existing debate/direct-execution path to the CLI mirror streaming approach, we needed to run both in parallel to evaluate CLI mirror’s real-world performance and catch any bugs. The CLI mirror approach provides 1.5s streaming updates with visible tool calls in Discord threads — a significantly […]

Summary Built and shipped a live audio PM interview practice tool at pezant.ca/interview/. Claude CLI (Max subscription) as interviewer, browser SpeechSynthesis for TTS with 50+ voice options, Web Speech API for STT with auto voice activity detection. Loads 7+ interview prep files as context. Structured scoring rubric (5 categories, advance decision). All loose ends resolved: […]