df4d8a3a19
Tool primes (tools-file, tools-reasoning, tools-memory, tools-subagent, tools-elevate) moved from preTurn to agentSpawn across all agent configs. Context is now injected once at session refresh points (start, post-clear, post-compaction) rather than repeated on every turn. Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com> |
||
|---|---|---|
| .. | ||
| README.md | ||
| coding.json | ||
| default.json | ||
| devops.json | ||
| greybeard.json | ||
| qa.json | ||
| reviewer.json | ||
| writer.json | ||
README.md
Agents
Each .json file defines an agent configuration. Agent configs are installed to ~/.config/ollie/agents/ and selected at runtime via /agent <name>.
default
Base agent with no specializations. Provides the standard ollie toolset and hook wiring. Use this as the starting point when creating a new agent — copy default.json and add a prompt and any additional configuration.
coding
Coding-focused agent. Extends default with agent instructions covering code modification discipline, security practices, and exploratory reading habits. Also enables the execute_elevated_bash trusted tool for operations requiring elevated privileges.
System prompt
The base system prompt is loaded from SYSTEM_PROMPT.md by Go at session creation — it is not assembled from hooks. Tool-specific prompts (e.g. tools-file, tools-memory) are injected via preTurn hooks using $OLLIE/x/prime <name>, which reads a file from p/ and writes it to stdout with environment variable substitution.
bd prime is appended in agentSpawn in both agents — it injects the active beads context if a database is found, and silently does nothing otherwise.
The fully assembled system prompt is visible at s/<id>/systemprompt.
Adding a new agent
Copy default.json, rename it, and adjust the preTurn hook array to include the prime calls you want. Add a prompt field for agent-specific instructions appended after the base system prompt. Hooks, trusted tools, and generation parameters can all be customized per agent.