fix(agents): disable aggressive default schedules in sample agent configs - #1207
Closed
Hypn0sis wants to merge 449 commits into
Closed
fix(agents): disable aggressive default schedules in sample agent configs#1207Hypn0sis wants to merge 449 commits into
Hypn0sis wants to merge 449 commits into
Conversation
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
(cherry picked from commit 7432086)
(cherry picked from commit 336240b)
The agent config reload logic was missing skills and mcp_servers from the change detection, so edits to these fields in agent.toml weren't being picked up when loading agents from SQLite. Added both fields to the comparison to ensure proper hot-reload.
The skills and mcp_servers fields must be at the top level of the agent.toml, not after [capabilities], due to TOML implicit table ordering rules.
… detection - Bump workspace version and Tauri config to 0.5.5 (fixes users stuck on 0.5.1) - Add ssrf_allowed_hosts config for self-hosted K8s environments (Jerry Jaz) - Raise Ollama discovered model defaults to 128K context / 16K output (Cureator) - Expand embedding auto-detection: OpenAI, Groq, Mistral, Together, Fireworks, Cohere, then local providers (Thunder Guardian) All tests passing. 9 files changed, 272 insertions.
feat: add SearXNG search provider with custom URL and JSON output
fix: Alpine Expression Error in settings page caused by x-show
Fix/agent skills reload
channels: add startup timeout for Telegram control-plane calls
feat(runtime): recover nested XML tool call parameters
Fix token estimation: include ToolUse arguments in text_length
Fix silent reinforcement: recognize [SILENT] token
Replace SHA256 password hashing with Argon2id for dashboard auth
feat(discord): smart auto-thread mode (true/false/smart)
fix: system prompt and identity handling, and config form hydration
…figs Commit 8f20cc2 ("chat agents", v0.6.9) introduced auto-spawning of all agents found in ~/.openfang/agents/ on boot (issue #1140). Three sample agents carried [schedule] sections that were silently ignored before auto-spawn existed: orchestrator: continuous every 120s -> ~30 LLM calls/hr ops: periodic every 5m -> ~12 LLM calls/hr health-tracker: periodic every 1h -> ~1 LLM call/hr Together these generated ~630 unexpected LLM calls/day and ~5 EUR/day in API costs for any user who installed the sample agents before enabling auto-spawn. Fix: remove active [schedule] sections from all three files and replace with commented-out examples at conservative intervals. Users who want autonomous mode can uncomment and tune the values explicitly. security-auditor is not changed — its proactive/event-driven schedule only fires on agent_spawned/terminated events, not on a timer.
This was referenced May 20, 2026
Contributor
Author
|
The Security Audit CI failure is unrelated to this PR's changes. Root cause: RUSTSEC-2026-0141 ( The Fix is in a separate PR (upgrading |
Hypn0sis
force-pushed
the
pr/fix-sample-agent-schedules
branch
from
July 3, 2026 07:51
5e1aea3 to
e5a0c46
Compare
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
20 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Fixes #1206.
What changed
Removed active
[schedule]sections from three sample agent configs that were generating unexpected LLM calls after the auto-spawn feature landed inefbefa1(v0.6.9):agents/orchestrator/agent.toml— wascontinuous { check_interval_secs = 120 }(~30 calls/hr)agents/ops/agent.toml— wasperiodic { cron = "every 5m" }(~12 calls/hr)agents/health-tracker/agent.toml— wasperiodic { cron = "every 1h" }(~1 call/hr)Each section is replaced with a commented-out example at a conservative interval so users can opt-in explicitly.
agents/security-auditor/agent.tomlis unchanged — itsproactiveschedule is event-driven, not timer-based.Why
Before
efbefa1, agents in~/.openfang/agents/were never auto-spawned on boot, so their[schedule]sections were inert. Afterefbefa1, they activate immediately on daemon restart, generating ~630 unexpected LLM calls/day and ~€5/day in API costs for users who have sample agents installed.