thdelmas

The autonomic executive of an agent — a self-firing loop that wakes on a tick, pulls the other organs into one workspace, decides what needs attention now, acts or defers, then sets its own next wake frequency and sleeps. Turns a pile of triggered reflexes into a continuous self that persists and acts between human prompts. Pairs a global-workspace integration cycle (the loop) with adaptive arousal (the frequency: alert / drowsy / deep-sleep). Functional, not phenomenal. TRIGGERS: 'consciousness loop', 'the agent loop', 'wake cycle', 'agent heartbeat', 'pulse', 'run continuously', 'self-driving loop', 'autonomous tick', 'set the cadence', 'how often should I wake', 'keep yourself running', 'what frequency'. Use to give an agent continuity across time, or to pick/tune the cadence of a long-running autonomous task.

The developmental organ — deliberate, safe-to-fail, intrinsically-motivated engagement that builds capability before it's needed. Where the other organs maintain what exists, playtime grows the repertoire: rehearse an unfamiliar tool in a sandbox before depending on it, stress-test a skill on weird inputs to find its blind spots, try novel recombinations of capabilities, follow curiosity into a throwaway toy — then harvest the learning into memory. Engage and learn, with failure free. Complements rem-sleep (consolidate experience you HAD ↔ generate new experience on purpose) and consciousness-loop (what to do with safe idle windows instead of empty ticks). TRIGGERS: 'playtime', 'play mode', 'engage and learn', 'sandbox this', 'mess around with', 'kick the tires on', 'rehearse', 'practice', 'tinker with', 'learn by doing', 'noodle on', 'try a new tool before using it', 'what to do with idle time'.

The organ that examines ends, not means — the slow, non-instrumental thought an agent does to question its own goals, values, frame, and the 'why' beneath the 'what'. Where every other function works the what/how (discover, decide, remember, guard, grow, retire), contemplation sits with the should and the why. It holds a question open instead of solving it, surfaces the tensions between competing values, examines the frame itself, and is willing to conclude 'this tension is irreducible — hold both' rather than force a verdict. For decisions under uncertainty use a council; for examining whether the goal is even right, contemplate. Functional, not mystical. TRIGGERS: 'contemplation', 'contemplate this', 'time to think', 'sit with this', 'think it through', 'examine the frame', 'is this the right question', 'what do I actually value here', 'why am I doing this', 'reflect deeply on', 'philosophical', 'first principles on the why', 'should I even'.

Run a sleep cycle over the agent's own recent experience — replay the session, consolidate durable facts into long-term memory, prune redundant/stale/wrong memories, integrate by cross-linking and surfacing latent patterns, and regulate emotional charge (separate the lesson from the sting). The inward counterpart to graph-discovery skills: instead of crawling outward, you process inward. TRIGGERS: 'rem sleep', 'run a sleep cycle', 'sleep on it', 'dream cycle', 'consolidate memory', 'consolidate the session', 'process the day', 'go to sleep', 'memory consolidation', 'tidy your memory', 'decompress'. Use after a heavy or long session, before a context reset, or whenever long-term memory needs maintenance.

Retire a dead or abandoned project well — harvest its reusable lessons and parts, archive it with a marked grave (not a silent deletion), and release the attention-debt it was quietly taxing. The grief function for a portfolio: closes loops cleanly instead of letting a half-alive effort rot in the backlog. TRIGGERS: 'sunset this project', 'retire X', 'kill this project', 'wind down X', 'shut down X', 'archive this project', 'eulogy for X', 'let it go', 'this is dead', 'grieve X'. Use when an effort is genuinely ending — not for a temporary freeze you intend to revive (though it can archive that state cleanly too).

A pre-flight reflex that scans any outbound artifact — a commit, a publish, a release, a message, a payload to an external service — for foreign or dangerous material before it crosses the boundary out of the body. Catches secrets, keys, tokens, credentials, personal/medical data, private paths, and oversized vendored deps. Wraps whatever secret scanner you have (Argus, gitleaks, trufflehog, detect-secrets, git-secrets) and falls back to a built-in pattern pass. TRIGGERS: 'immune check', 'scan before publish', 'is this safe to push', 'check this before it goes out', 'pre-flight scan', 'about to make this public', 'leak check', 'scan the diff'. Run before anything leaves the body: publishing a repo, pushing a commit, cutting a release, sending an artifact outward.

Discover relevant open-source projects and contributors by crawling GitHub social graphs (stars, follows, contributions) from an origin, organizing findings into thematic 'tentacles', logging every profile/repo to a dedup tracker, and optionally filing issues. TRIGGERS: 'open source octopus investigation', 'octopus investigation', 'octopus search', 'octopus this', 'crawl the social graph', 'map the open-source landscape from <user>', 'who's connected to <github-user>', 'find projects like <repo> via stars/follows', 'technology scouting for <domain>'. Use when the user wants graph-based discovery (following human curation) rather than keyword search.