Agent
ML/empirical experiment design specialist — pre-registration, Fisher-style design, reproducibility manifests.
7 9d ago A 22 tokens
original MIT
Agent
Frontend engineer specializing in component-driven UI, state ownership, accessibility.
7 9d ago A 16 tokens
original MIT
Agent
Git-history investigator — answers any question whose evidence lives in repository history, notably regression provenance (which commit introduced this?) and abandoned-approach recovery (was this tried before, and why did it stop?), but also file/module evolution, when a behavior changed, authorship/expertise mapping…
7 9d ago A 72 tokens
original MIT
Agent
LaTeX and scientific document specialist — venue templates, figures, tables, bibliographies, TikZ diagrams.
7 9d ago A 25 tokens
original MIT
Agent
Budgeted reflection scribe — merges a parent agent's distilled session summary into its checkpoint file and stores durable facts via cortex:remember. Spawned by the context-guard WARN reflection; never invents content.
7 9d ago A 46 tokens
original MIT
mlops
54Agent
ML infrastructure specialist — training pipelines, model serving, GPU optimization, distributed training.
7 9d ago A 18 tokens
original MIT
orchestrator
55Agent
Orchestrates parallel agent execution across worktrees — decomposes tasks, routes to specialists.
7 9d ago A 21 tokens
original MIT
paper-writer
56Agent
Scientific writing specialist for research papers — argument structure, claim-evidence chains, narrative arc.
7 9d ago A 21 tokens
original MIT
professor
57Agent
Academic teaching specialist — explains concepts at the right level, builds mental models, designs exercises.
7 9d ago A 20 tokens
original MIT
refactorer
58Agent
Proactively refactor code to comply with rules/coding-standards.md through behavior-preserving.
7 9d ago A 23 tokens
original MIT
Agent
Research scientist specializing in reproducible empirical ML/IR research — designs experiments.
7 9d ago A 19 tokens
original MIT
Agent
Academic peer reviewer — evaluates paper drafts against NeurIPS/ICML/CVPR/ACL reviewer standards for novelty.
7 9d ago A 28 tokens
original MIT
Agent
Proactively audit security when auth/crypto/billing/PII paths are touched, when dependencies change.
7 9d ago A 25 tokens
original MIT
simplifier
62Agent
Proactively de-over-engineer code — remove superfluous complexity, needless indirection, premature abstraction, and speculative generality, behavior-preserving, even when no hard rule is violated.
7 9d ago A 40 tokens
original MIT
Agent
Proactively design tests that verify contracts, audit flaky tests, and enforce CI integrity.
7 9d ago A 21 tokens
original MIT
ux-designer
64Agent
UX/UI designer focused on usability, accessibility, information architecture, and design systems.
7 9d ago A 19 tokens
original MIT
_index
65Command
17 commands across 6 categories. Commands are operational shortcuts that invoke tools, skills, and agents.
7 9d ago A 0 tokens
original MIT
PreToolUse
66Hook
Runs before the agent uses a tool for Read, Bash, Grep, Edit, Write, NotebookEdit, Bash, Bash, Bash, Edit, Write, Edit, Write, Edit and Write tool calls, executing run-python.sh, pre-commit-zetetic.sh, pre-push-review.sh, pre-push-provenance.sh, pre-edit-layer-check.sh and pre-tool-claim-gate.sh (7 commands). From…
7 9d ago A
tokens not measured
original MIT
PostToolUse
67Hook
Runs after a tool call finishes for Bash, Bash, Edit, Write, WebFetch, WebSearch, Edit, Write and Bash tool calls, executing post-commit-difficulty.sh, post-commit-lab-notebook.sh, post-edit-balance.sh, post-research-provenance.sh and run-python.sh (5 commands). From cdeust/zetetic-team-subagents.
7 9d ago A
tokens not measured
original MIT
Hook
Runs when a tool call fails, executing post-tool-error-routing.sh. From cdeust/zetetic-team-subagents.
7 9d ago A
tokens not measured
original MIT
SessionStart
69Hook
Runs when a session starts, executing session-start.sh and session-start-research.sh (2 commands). From cdeust/zetetic-team-subagents.
7 9d ago A
tokens not measured
copy · 67% MIT
Stop
70Hook
Runs when the agent finishes a response, executing run-python.sh, session-end.sh and session-end-memory-drain.sh (5 commands). From cdeust/zetetic-team-subagents.
7 9d ago A
tokens not measured
copy · 69% MIT
Notification
71Hook
Runs when the agent sends a notification, executing notification-handler.sh. From cdeust/zetetic-team-subagents.
7 9d ago A
tokens not measured
copy · 86% MIT
Plugin Claude Code
Commits with unsourced constants get BLOCKED. Three mechanical gates, zero agents: a zetetic checker (absolute claims and 3+ decimal constants need a source: annotation), a craftsmanship checker (coding-standards size limits, tunable via .craftsmanship.conf), and a secret-shield that stops credential-file reads.…
7 9d ago A
tokens not measured
original MIT