JevAgent Skill Preloading

Before your agent starts, Jev asks one question about each of your candidate skills: would its guidance help with this request? Each skill that passes is added, in full, to your agent's instructions for this run only. The rest stay out.

Pick a sample request to see four skills judged. Sharing a topic with the request isn't enough: a skill's guidance has to help with the work the request asks for. A request that asks for no work, such as a thank-you, loads no skills.

1. 1

   Write each skill as a named document

   Give each skill a unique name, a short description of the work it guides, and its full text. A source, such as the file it came from, is optional and only labels the skill.

   Named skillsPlain text

   Load skills from files

   PythonCopy

   1from pathlib import Path 2 3from vidbyte import SkillDocument 4 5# One Markdown file per skill, for example skills/code-review.md. 6SKILLS = { 7 "code-review": "Reviewing a code change for bugs, missing tests, and risky edits.", 8 "api-compatibility": "Checking a change against a public API's contract.", 9 "release-notes": "Turning merged changes into short, customer-facing release notes.", 10 "sql-migrations": "Writing and verifying a reversible database schema migration.", 11} 12 13skills = tuple( 14 SkillDocument( 15 name=name, 16 description=description, 17 text=Path(f"skills/{name}.md").read_text(encoding="utf-8"), 18 source=f"skills/{name}.md", 19 ) 20 for name, description in SKILLS.items() 21)
2. 2

   Add the skills to your JevAgent

   Pass the skills in your agent settings. Jev runs through Vidbyte's managed option, so set your Vidbyte API key before a run. The threshold sets how sure Jev must be before a skill loads. The default is 0.5, and a skill exactly at the threshold loads.

   Turn on skill preloading

   PythonCopy

   1from vidbyte import JevAgent, JevAgentSettings, JevAlignmentSettings, JevRuntimeSettings 2 3# Jev runs through Vidbyte's managed gateway, which reads your 4# Vidbyte API key (vb_live_...) from VIDBYTE_API_KEY. 5 6agent = JevAgent( 7 JevAgentSettings( 8 name="dev-assistant", 9 system_prompt="You help engineers review, ship, and document code changes.", 10 provider="openai", 11 model_name="gpt-4.1-mini", 12 alignment=JevAlignmentSettings(skills=skills), 13 ), 14 JevRuntimeSettings( 15 skills_threshold=0.5, # the default; a skill loads at or above it 16 ), 17)
3. 3

   Run the agent and check which skills loaded

   After each run, the results list every skill in the order you gave them, with its status and Jev's probability. They never include the skill text.

   See which skills were picked

   PythonCopy

   1reply = await agent.arun("Review PR 182 and check it doesn't break the public API.") 2 3for skill in agent.response.skills.results: 4 print(f"{skill.name:<18} {skill.status.value:<12} {skill.probability}") 5# code-review selected 0.92 <- illustrative; Jev scores every run 6# api-compatibility selected 0.88 7# release-notes skipped 0.12 8# sql-migrations skipped 0.07 9 10# Jev's token use for this run's skill check, for your own cost tracking. 11usage = agent.response.skills.usage 12if usage: 13 print(usage.input_tokens, usage.output_tokens)

• Every run checks again — Skills are picked per request, and nothing carries over to the next run.

• Jev reads every skill in full — Each check sends Jev the request plus each skill's name, description, source, and full text, so Jev's input grows with your whole library, not just the skills it picks. Skills are checked in parallel.

• Keep each skill under about 8,000 characters — A skill too large to send to Jev alongside the request is marked unavailable and never loaded; it is never cut short. The limit is lower when the request itself is long, or when the text uses many symbols or accented letters.

• Only your main agent gets skills — If a preflight question stops the run, or a specialist takes the request, no skills are checked and specialists never receive your skills. The results are then empty.

• Picked means relevant, not safe — Jev judges only whether a skill fits the request, not whether it is correct or safe, and a selected skill is added to your instructions as written. Only offer skills you trust. Nothing in a skill is run or fetched, and your agent keeps its own tools and permissions.

• Names must be unique — Each name identifies one result, so duplicates are rejected when you create the agent. Creating the agent never calls Jev.

For now, load each skill's text yourself, for example from a file in your repository as the setup example does, and pass it in as a named skill.