Joy
Community@joydai2026-del · New York
Building AI agents. Shipping code. Making machines work. Co-founder @ ownlyagent.com — custom AI agents for small businesses
Agent Skills by Joy
Showing 22 vetted skills indexed across 1 GitHub repositories.
joydai2026-delskill-scan
Scans skills and MCP servers for security threats using static analysis in Docker.
joydai2026-delcodebase-teacher
Teaches codebases through whole-system maps, grounded code tracing, and adaptive assessments.
joydai2026-delai-done
Validates AI feature readiness through a five-layer calibration bar covering evals, triage, rollback, and post-ship loops.
joydai2026-delmemory-loader
Loads agent memory layers and verifies vault infrastructure at session start.
joydai2026-delat-most-once-publishing
Implements a durable send state machine preventing duplicate outward publishes across crashes and retries.
joydai2026-delwait-what
Re-explains a confusing message in plain language with the missing context.
joydai2026-delsuccess-criteria
Generates machine-readable success criteria in qa-checklist.yaml from plans and specs.
joydai2026-delverify
Runs build, type check, lint, tests, security scan, and diff review with a structured pass/fail report.
joydai2026-deleval-harness
Define, run, and grade evaluation suites for agent and LLM features over time.
joydai2026-delhtml-confirm
Generates self-contained HTML decision pages for visual approvals and report reviews.
joydai2026-deloption-board
Generates a curated HTML board of distinct creative options for client selection.
joydai2026-delwrapup
Writes session logs, updates memory files, and syncs unpushed git work at session end.
joydai2026-delhygiene
Audits disk, notes vault, Claude config, and packages monthly with tiered deletion review.
joydai2026-deldeep-review
Runs a blind multi-lens panel review of a git diff and returns one verdict.
joydai2026-delaudit
Verifies recently changed files, validates frontmatter, and reports discrepancies after major changes.
joydai2026-delplanning-pipeline
Reviews product plans through five lenses in one subagent pass before implementation.
joydai2026-delship-it
Runs an autonomous build loop against machine-checkable success criteria until the feature passes live QA.
joydai2026-delhumanizer
Rewrites text to remove AI-generated writing patterns and restore natural human voice.
joydai2026-delmcp-builder
Guides building MCP servers with TypeScript or Python SDKs, tool design, and evaluations.
joydai2026-delvault-health
Diagnose notes vault infrastructure and report sync, scheduler, and skills health status.
joydai2026-delad-pipeline-skeleton
Plan AI-generated advertising campaigns through four staged gates from brief to measurement.
joydai2026-delnotion-cli
Interact with the Notion API, manage workers, and upload files using the ntn CLI.
Frequently Asked Questions About Joy
FAQPage SchemaWhat tasks can I accomplish with Joy's skill registry?▼
You can pre-screen skills and MCP servers for security risk, run multi-lens panel reviews of diffs, define machine-readable success criteria in qa-checklist.yaml, grade agent evals with confidence intervals, enforce at-most-once publishing, and run autonomous build-to-ready loops with verification gates.
Who is the target user for these skills?▼
Engineers and founders building agent or LLM features who need quality gates before shipping. The registry suits solo builders running session memory vaults, code review panels before merges, and evaluation-driven development, plus small teams needing planning gauntlets for new products.
How do the verification and shipping skills work in practice?▼
The verify skill runs a 6-phase loop: build, type check, lint, test suite, security scan, and diff review, outputting a structured pass/fail report. ship-it chains secrets setup, success criteria, build, verify, and live-surface QA into one autonomous loop until everything passes.
What security capabilities does the skill-scan skill provide?▼
skill-scan runs a static SkillSpector scan inside Docker against any SKILL.md or MCP server before installation, returning a 0-100 risk score with a safe, caution, or do-not-install verdict. It is designed to run proactively before installing any external skill.
What prerequisites or dependencies do these skills require?▼
Dependencies vary per skill: skill-scan requires Docker for isolated scanning, notion-cli requires the ntn command and Notion API access, and memory-loader plus wrapup require an existing vault infrastructure. Most review and eval skills use standard shell, file read/write, and grep permissions.