Skills are not available in Agent.
Nobody has claimed this yet.
- Dominant language
- Go
- Stars
- 49.2k
- Forks
- 4.5k
- Avg merge
- 1d 3m
- Merged PRs (30d)
- 239
Description
LocalAI version:
Latest (4.2.6)
Linux/Arch/Docker
Environment, CPU architecture, OS, and Version:
Linux mandy 7.0.3-arch1-2 #1 SMP PREEMPT_DYNAMIC Fri, 01 May 2026 15:49:22 +0000 x86_64 GNU/Linux
Describe the bug
Agent reports no skills available, after making skills and exposing them to an agent.
Since this is listed as a working subsystem in the LocalAGI project, my best guess is that the lack of skill exposure is some sort of configuration issue at the LocalAI level?
To Reproduce
- Turn on agent pool
- create an agent
- create a skill (in my case "talk-like-pirate" as a test)
- Enable Skills for that agent
- Optional: select some skills
- Ask agent to "list skills"
Observe that skills_list is always empty [].
Expected behavior
Agent should find and be able to use skills. The UI says that they will be "exposed as MCPs to the agent".
Logs
I won't spam you with logs unless you want...possibly this is simply a known issue.
Additional context
Does this mean I have to add the skills as an MCP manually? It is quite unclear what I would enter into the Agent MCP configuration. Could be that I have located another "documentation enhancement request"?
Thanks much.
Contributor guide
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
Research direction
Reproduce the issue through the agent pool, agent skill settings, and the skills_list response, using the reported Linux/Arch/Docker setup if available. Trace how enabled skills are exposed as MCPs to an agent and compare that path with the empty skills_list result. Done means an agent can list and use its enabled skills, with coverage for the reported setup.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- go
- Domain
- ai, backend-api-design
- Issue type
- Bug
- Difficulty
- 3/5
- Estimated time
- 1-2 days
- Activity status
- Active
- Clarity
- Mostly clear
- Newbie friendliness
- 55/100