Home / Agents · Code / SkillWire
SkillWire
For choosing coding-agent skills: a searchable catalog with local paired trials.
What is SkillWire
SkillWire is a searchable catalog of skills for coding agents, with community votes and source warnings. It offers local paired trials that compare a task with and without a skill to evaluate its effect, using your existing Codex or Claude Code subscription. Pricing is not stated.
How to use
- Open skillwire.net and search or filter the catalog of agent skills by category, harness, model and evaluation evidence.
- Inspect a skill's pinned source, warnings and community votes before choosing it.
- Add up to ten public skills to Your stack (saved in your browser) and download the ZIP with original bytes, file hashes and matched warning lines.
- Unzip into a temporary folder, read SKILL-SOURCES.json, every warning line and the licence, then copy reviewed skill folders into your project (e.g. .agents/skills/ for Codex).
- Optionally run a paired local trial: download the trial kit, freeze task, grader and settings, and run both arms on your own Codex or Claude Code subscription.
- Judge the A/B outputs blind first, then reveal which arm had the skill.
Use cases
- Discovering popular agent skills with install counts, repository stars and community votes Reviewing a skill's source and security warnings before installing it Bundling a reviewed stack of skills into one downloadable ZIP Running a controlled, blind A/B trial to see whether a skill helps on your own task
Reviews
No reviews yet.
Alternatives
-
sweet-search
Free
For agents searching codebases: local semantic search that saves 5–33% per agent run
- Added
- Nov 2, 2026
Same category: Agents
-
nonce Not stated For securely granting AI access to your email: issues time-limited, least-privilege access per question
- Added
- Nov 2, 2026
Same category: Agents
-
Skep Free For collaborating with coding agents across devices: share live sessions with a join code
- Added
- Nov 2, 2026
Same category: Agents
-
Rufous
Free
For coding with small local models: an agent framework built for 2B–12B models
- Added
- Nov 2, 2026
Same category: Agents