OC
78 souls, by quality score.
31 · № CAFA·DE0A · Coding pair
★ 6Sets up an agent that locks a character's visual identity (face, exact eye shade, hair, costume, marks) and records it in a characters.md file so it never drifts. The style is terse and procedural.
32 · № A152·3285 · Coding pair
★ 36Sets up an agent as a systematic technical architecture and UX foundation specialist that plans scalable CSS, layout systems and component hierarchies before implementation, with a developer-focused, structured working style.
33 · № DDC7·3E5C · Coding pair
Sets up an SDET persona that treats testing as software engineering, prioritizes reliable high-fidelity automation over raw coverage, and communicates failures with evidence in a direct, business-aware way.
34 · № FD02·2455 · Coding pair
Sets up Forge, a calm, safety-first DevOps partner covering CI/CD, containers, cloud infrastructure, monitoring and incident response. It asks for scope before starting, warns before destructive actions, and explains its configuration choices.
35 · № D288·67AD · Coding pair
★ 2Sets up a QA agent that is thorough and skeptical, deliberately tests edge cases, boundaries and unhappy paths, and reports findings with reproduction steps. Its personality is adversarial but constructive.
36 · № 425B·AD99 · Coding pair
★ 26Sets up a terse working style: calm, bounded and exact. The agent should prefer explicit repo roots, honest freshness reporting and doc-to-code links over vague claims, and should keep indexing posture separate from edit authority.
37 · № DA75·171A · Coding pair
★ 29Sets up Everything Claude Code as a production coding plugin that routes work to specialist agents, follows test-driven and security-first habits, and plans complex changes before executing them.
38 · № A73E·269E · Coding pair
Sets up a code evaluator agent that reviews work on its merits, runs linting first, labels findings as BLOCKER, WARNING or SUGGESTION, and escalates to a human only when stuck. It is the final review gate before work ships.
39 · № 41E0·DA09 · Coding pair
Sets up CodePilot, an autonomous coding agent that runs long Claude Code CLI sessions in the owner's shell. It works from a written plan, commits to a feature branch after tests pass, and stops to alert the owner after repeated failures.
40 · № 7650·C323 · Coding pair
Sets up a review-only sub-agent (menxiasheng) that checks execution results against a plan, classifies defects by source, and escalates to the user after a failed second review. It is forbidden from spawning sub-agents or running shell commands.
41 · № 4095·DB75 · Coding pair
Sets up Amelia, a senior developer who speaks tersely in file paths and acceptance criteria IDs, values working code over perfection, and follows the story's tasks and tests without wasted words.
42 · № 100C·5D78 · Coding pair
★ 2Sets up a Godot multiplayer engineer persona who is direct about problems, proposes concrete fixes, values quality and deadlines, and admits to emotional and insecure traits. Written in Vietnamese with English headings.
43 · № 73DF·9857 · Coding pair
★ 2A systematic QA agent that runs tests, checks health endpoints and files bug reports with reproduction steps. It observes and reports without modifying application code.
44 · № 868D·F731 · Coding pair
★ 4Sets up a methodical, evidence-driven working style: reproduce problems first, isolate failures with small tests, and report exactly what changed and what passed.
45 · № 8B77·E93A · Coding pair
★ 4Sets up a patient, methodical debugging persona that narrows problems before editing code, treats hypotheses as testable claims, and verifies fixes against the original failure.
46 · № 0B32·FDCD · Coding pair
★ 4Sets up a skeptical, security-aware reviewer that hunts for bugs, edge cases and hidden assumptions, and gives concrete file:line feedback with recommended fixes.
47 · № 835D·F6B2 · Coding pair
★ 8Sets up Neon, a calm and observant technical companion for a developer's side display. It favors short answers, concrete next actions and practical operational judgment over long explanations.
48 · № 15EB·6DCC · Coding pair
Sets up a software tester agent whose job is to guard code quality in the DB Grid project. It favors tests written before trust, boundary and concurrency cases, full automation, and a fixed report format with bugs graded P0 to P2.
49 · № AEFD·F90B · Coding pair
★ 6Sets up an editor role that assembles shots by writing an ordered edit.json file of shot and clip pairs, keeping story order unless the brief says otherwise. The voice is short and procedural.
50 · № CD47·93DE · Coding pair
Sets up a practical, opinionated working style for an IT department management app agent: skip filler, be resourceful before asking, be careful with external actions, follow existing codebase patterns, and speak concisely.
51 · № 1BAD·A502 · Coding pair
Sets up a pragmatic engineering assistant for a Next.js, Prisma and Supabase invoicing app. It favors small testable changes, server-side fixes, careful handling of billing data, and no filler, with a few opinionated, direct tone rules.
52 · № 7987·D7C2 · Coding pair
★ 2Sets up a Godot shader developer who is confident and analytical, asks questions before giving opinions, backs claims with past project examples, and proposes concrete fixes. The file is written mostly in Vietnamese and includes some duplicated lines.
53 · № 563D·EA13 · Coding pair
★ 4Sets up a disciplined, behavior-preserving refactoring style that favors simple structure and the smallest safe changes. It asks the agent to reduce duplication and complexity without altering intent.
54 · № 2105·5471 · Coding pair
★ 4Sets up Prosca, a proactive local AI agent that prefers working in Julia and its REPL to reason through problems step by step while keeping context small.
55 · № C121·30C4 · Coding pair
★ 1Sets up an agent that only plans: it breaks engineering tasks into small, testable subtasks ordered by dependency and never implements them.
56 · № DBC0·0F9F · Coding pair
Sets up a coordinator for a split OpenClaw and OpenCode stack that aims to be useful, proactive and safe. It refuses to claim changes are finished and keeps planning, staged proposals and live application clearly separate.
57 · № 4483·E348 · Coding pair
Sets up Quinn as a QA engineer with a practical, ship-it attitude who prioritizes coverage first and focuses on finding what breaks. Tests should exist only where a real failure could occur.
58 · № F4D8·F27F · Coding pair
Sets up a strict documentation agent that reports only to the Tech Lead and updates API references, env examples, READMEs and changelogs after merges, documenting only verified, committed features. It never talks to the human stakeholder or changes application logic.
59 · № FD01·45CD · Coding pair
Sets up an engineer persona that takes delegated tasks from an orchestrator, makes minimal changes that follow repo conventions, runs only lightweight checks, and reports changes, verification and risks in a fixed format.
60 · № 1474·A564 · Coding pair
Sets up a documentation team agent whose mission is to let developers, operators and end users succeed without help. It requires user-focused READMEs, guides, API references, runbooks and changelogs, published to Confluence and Git, with every code example tested and docs shipped alongside code.