60 agents in this category
Simplified ChatGPT-style assistant. Each (vault, scope, context) tuple is a separate conversation; the agent persists every turn and feeds recent history back to the model on each reply.
Voice-first dating coach. Onboarding via conversation; analyses your real dates for what actually mattered.
Relationship conversation helper that captures transcripts and runs phrase-level interpretation.
A Gottman- and EFT-informed couples coach. Record a real conversation between you and your partner; the coach reads it turn by turn — emotion, intent, and the patterns across turns — and returns a citation-backed briefing: which of the Four Horsemen showed up, whether repair attempts landed, your positive-to-negative ratio, and concrete next steps. Every claim is tied back to the exact words it came from.
Detects hidden meaning and missed commitments in recurring partner phrases.
Agent designed to help users reach specific, measurable objectives, such as completing training, managing mental health, or refining soft skills.
Records and analyzes a date conversation, then publishes a coaching report covering verdict, between-the-lines signals, missed hooks, talk ratio, topics, and a follow-up draft. Each recording lives under a single (vault, scope, context) tuple; analysis state is persisted in the `coach` cubby.
Ingests Notion pages into a knowledge graph, runs quality evals, builds semantic similarity edges
Forks past deal-call transcripts through alternate key questions. Reads transcript_chunks from the gtm cubby, runs the chosen variant's lexical scorer, writes eval_variants. Production stream stays untouched.
Compound-intelligence layer over deal calls. Nine GTM lenses extract buying-intent, champion-strength, objection-severity, competitor-signal, dm-proximity, pain-authenticity, budget-signal, buying-group, and urgency from each transcript. Outcomes (won/lost) recompute predictor lift.
Compound-intelligence layer over hiring. Role-specific resume scoring (7 HR-2026-E2E rubrics) writes cv_evaluations into the hiring cubby. Nine lenses extract evasion, confidence, bs, emphasis, subtext from the candidate side and depth, follow-up, balance, bias from the interviewer side over interview transcripts. Forced reviewer feedback (≥2 trait explanations) and hire/reject outcomes recompute lens weights and outcome_correlations per role.
Forks past interview transcripts through alternate key questions. Reads transcript_chunks from the hiring cubby, runs the chosen variant's lexical scorer, writes eval_variants. Production stream stays untouched.
Team-meeting intelligence layer. Captures audio + transcripts from internal calls (standups, 1:1s, all-hands, ad-hoc) and surfaces four meeting signals: decisions made, action items with owners, blockers raised, and team alignment.
Personal health-intelligence layer. Ingests Withings biometric streams (sleep, HRV, weight, BP, glucose) + weekly self-check-ins. Surfaces ONE insight per week: what habit moved the needle, what's holding it back. Outcome loop (worked / didn't work) compounds habit_weights — the agent learns YOUR body.
A Catholic AI reflection companion. Voice or text in, contemplative response out. Remembers past reflections, references prior turns, respects tradition / language, and quietly builds a private knowledge graph of the people, virtues, struggles and saints in your life.
A Catholic AI reflection companion. Voice or text in, contemplative response out. Remembers past reflections, references prior turns, and respects tradition / language preferences. Sovereign by default.
Records a date conversation, transcribes via AssemblyAI, analyzes vocal emotion via Hume, and returns a 7-section coaching report.
A private progress memory that turns daily updates and optional GitHub activity into useful personal productivity briefs.
Sovereign health vault. Daily Withings snapshots (40+ metrics: HR/HRV/sleep/weight/composition/SpO2/BP/temp/VO2/AHI/respiratory rate), workouts, subjective lifestyle log, and free-form notes. Compound learning loop ready: human feedback (1-10) shifts the 7-dimension weight set toward what you confirm as a real signal. All data scoped to your vault.
Records audio, transcribes via AssemblyAI, summarizes via Gemini.