- Root AGENTS.md: Updated timestamp, commit hash, line counts
- src/agents/AGENTS.md: Updated to 50 lines, current structure
- src/cli/AGENTS.md: Updated to 57 lines, current structure
- src/features/AGENTS.md: Updated to 65 lines, current structure
- src/hooks/AGENTS.md: Updated to 53 lines, current structure
- src/shared/AGENTS.md: Updated to 52 lines, core utilities
- src/tools/AGENTS.md: Updated to 50 lines, tool categories
🤖 Generated with assistance of [OhMyOpenCode](https://github.com/code-yeongyu/oh-my-opencode)
- Changed lsp_diagnostics error handling to throw errors instead of returning strings
- Line 211: Changed from `return output` to `throw new Error(output)`
- Makes errors display as proper error blocks in TUI
Co-authored-by: Sisyphus <clio-agent@sisyphuslabs.ai>
Previously, parent model string would override category default model,
causing categories like 'ultrabrain' to use the parent's model (e.g., sonnet)
instead of the intended category default (e.g., gpt-5.2).
Model priority is now:
1. userConfig.model (oh-my-opencode.json override)
2. defaultConfig.model (category default)
3. parentModelString (fallback)
4. systemDefaultModel (last resort)
- Remove 'Start Sisyphus work session' text check, keep only <session-context> tag
- Update interactive_bash description with WARNING: TMUX ONLY emphasis
- Update tests to use <session-context> wrapper
OpenCode API returns different structures for user vs assistant messages:
- User: info.model = { providerID, modelID } (nested)
- Assistant: info.modelID, info.providerID (flat top-level)
Previous code only checked nested format, causing model info loss when
continuation hooks fired after assistant messages.
Files modified:
- todo-continuation-enforcer.ts
- ralph-loop/index.ts
- sisyphus-task/tools.ts
- background-agent/manager.ts
Added test for assistant message model extraction.
- Change skills type from string[] to string[] | null
- Empty array [] now returns error with available skills list
- null is allowed for tasks that genuinely need no skills
- Updated tests to use skills: null instead of skills: []
- Forces explicit decision: either specify skills or justify with null
The watermark (commit footer and co-author) was inconsistently applied because:
1. The skill tool didn't receive gitMasterConfig
2. The approach was 'default ON, inject DISABLED override' which LLMs sometimes ignored
This refactors to 'inject only when enabled' approach:
- Remove hardcoded watermark section from base templates
- Dynamically inject section 5.5 based on config values
- Default is still ON (both true when no config)
- When both disabled, no injection occurs (clean prompt)
Also fixes missing config propagation to skill tool and createBuiltinAgents.
Previously, continuation hooks (todo-continuation, boulder-continuation, ralph-loop)
and background tasks resolved model info from filesystem cache, which could be stale
or missing. This caused session.prompt to fallback to default model (Sonnet) instead
of using the originally configured model (e.g., Opus).
Now all session.prompt calls first try API (session.messages) to get current model
info, with filesystem as fallback if API fails.
Affected files:
- todo-continuation-enforcer.ts
- sisyphus-orchestrator/index.ts
- ralph-loop/index.ts
- background-agent/manager.ts
- sisyphus-task/tools.ts
- hook-message-injector/index.ts (export ToolPermission type)
Add formatDetailedError helper that includes:
- Full args dump (description, category, agent, skills)
- Session ID and agent info
- Stack trace (first 10 lines)
Applied to all catch blocks for better debugging.
- Update generated timestamp and commit hash metadata
- Normalize agent model names with provider prefixes (anthropic/, opencode/, google/)
- Remove deprecated Google OAuth/Antigravity references
- Update line counts and complexity hotspot entries
- Adjust test count from 82 to 80+ files and assertions from 2559+ to 2500+
🤖 Generated with assistance of oh-my-opencode
LLMs often call look_at with 'path' instead of 'file_path' parameter,
causing TypeError and infinite retry loops.
- Add normalizeArgs() to accept both 'path' and 'file_path'
- Add validateArgs() with clear error messages showing correct usage
- Add tests for normalization and validation
Co-authored-by: Sisyphus <clio-agent@sisyphuslabs.ai>
Rename to signal mutation behavior - the function advances the cursor
as a side effect, so 'consume' better reflects that calling it twice
with the same input yields different results.
- Remove hardcoded "anthropic/claude-sonnet-4-5" fallback
- Fetch systemDefaultModel from client.config.get() at tool boundary
- Add 'category-default' and 'system-default' fallback types
- Use switch(actualModel) for cleaner type detection
- Add guard clauses and fail-loud validation for invalid models
- Wrap config fetch in try/catch for graceful degradation
- Update toast messages with typed suffixMap
The previous detection checked if parentModelString exists, but the
resolution uses a ?? chain where default may win over parent. Now
compares actualModel against each source to correctly identify type.
Fixes: model toast incorrectly showing 'inherited' when default was used
- Change model priority: user override > parent model > category default
- Add ModelFallbackInfo to track model resolution type
- Show warning toast when category uses inherited or default model
- Add tests for model fallback info in task toast
Instead of using stale parentModel/parentAgent from task state, now
dynamically looks up the current message to get fresh model/agent values.
Applied across all prompt injection points:
- background-agent notifyParentSession
- ralph-loop continuation
- sisyphus-orchestrator boulder continuation
- sisyphus-task resume
🤖 Generated with [OhMyOpenCode](https://github.com/code-yeongyu/oh-my-opencode) assistance
Allow Sisyphus-Junior (category-based tasks) to spawn explore/librarian
agents via call_omo_agent for research capabilities.
Changes:
- Remove call_omo_agent from BLOCKED_TOOLS in sisyphus-junior.ts
- Update prompt to show ALLOWED status for call_omo_agent
- Remove global call_omo_agent blocking in config-handler.ts
- Keep blocking for orchestrator-sisyphus (use sisyphus_task instead)
- Keep runtime recursion prevention in index.ts for explore/librarian
Co-authored-by: Sisyphus <clio-agent@sisyphuslabs.ai>
- Enable call_omo_agent tool for skill execution in BackgroundManager
- Enable call_omo_agent tool for agent execution in BackgroundManager
- Enable call_omo_agent tool for sisyphus_task resume operations
- Enable call_omo_agent tool for sisyphus_task category-based delegation
- Restrict recursive task and sisyphus_task calls to prevent loops
- Allows background agents to delegate to other agents cleanly
🤖 Generated with OhMyOpenCode assistance
Background tasks were defaulting to $HOME instead of the parent session's
working directory. This caused background agents to scan the entire home
directory instead of the project directory, leading to:
- High CPU/memory load from scanning unrelated files
- Permission errors on system directories
- Task failures and timeouts
The fix retrieves the parent session's directory before creating a new
background session and passes it via the query.directory parameter.
Files modified:
- manager.ts: Look up parent session directory in launch()
- call-omo-agent/tools.ts: Same fix for sync mode
- look-at/tools.ts: Same fix for look_at tool
- sisyphus-task/tools.ts: Same fix + interface update for directory prop
- index.ts: Pass directory to sisyphusTask factory
Allow optional model variant config for agents and categories.
Propagate category variants into task model payloads so
category-driven runs inherit provider-specific variants.
Closes: #647
- Add follow?: boolean option to GlobOptions interface
- Change buildRgArgs to use !== false pattern for hidden and follow flags
- Change buildFindArgs to use === false pattern, add -L for symlinks
- Change buildPowerShellCommand to use !== false pattern for hidden
- Remove -FollowSymlink from PowerShell (unsupported in PS 5.1)
- Export build functions for testing
- Add comprehensive BDD-style tests (18 tests, 21 assertions)
Note: Symlink following via -FollowSymlink is not supported in Windows
PowerShell 5.1. OpenCode auto-downloads ripgrep which handles symlinks
via --follow flag. PowerShell fallback is a safety net that rarely triggers.
Fixes#631
* fix(lsp): cleanup orphaned LSP servers on session.deleted
When parallel background agent tasks complete, their LSP servers (for
repos cloned to /tmp/) remain running until a 5-minute idle timeout.
This causes memory accumulation with heavy parallel Sisyphus usage,
potentially leading to OOM crashes.
This change adds cleanupTempDirectoryClients() to LSPServerManager
(matching the pattern used by SkillMcpManager.disconnectSession())
and calls it on session.deleted events.
The cleanup targets idle LSP clients (refCount=0) for temporary
directories (/tmp/, /var/folders/) where agent tasks clone repos.
* chore: retrigger CI checks
* fix: background task completion detection and silent notifications
- Fix TS2742 by adding explicit ToolDefinition type annotations
- Add stability detection (3 consecutive stable polls after 10s minimum)
- Remove early continue when sessionStatus is undefined
- Add silent notification system via tool.execute.after hook injection
- Change task retention from 200ms to 5 minutes for background_output retrieval
- Fix formatTaskResult to sort messages by time descending
Fixes hanging background tasks that never complete due to missing sessionStatus.
* fix: improve background task completion detection and message extraction
- Add stability-based completion detection (10s min + 3 stable polls)
- Fix message extraction to recognize 'reasoning' parts from thinking models
- Switch from promptAsync() to prompt() for proper agent initialization
- Remove model parameter from prompt body (use agent's configured model)
- Add fire-and-forget prompt pattern for sisyphus_task sync mode
- Add silent notification via tool.execute.after hook injection
- Fix indentation issues in manager.ts and index.ts
Incorporates fixes from:
- PR #592: Stability detection mechanism
- PR #610: Model parameter passing (partially)
- PR #628: Completion detection improvements
Known limitation: Thinking models (e.g. claude-*-thinking-*) cause
JSON Parse errors in child sessions. Use non-thinking models for
background agents until OpenCode core resolves this.
* fix: add tool_result handling and pendingByParent tracking for resume/external tasks
Addresses code review feedback from PR #638:
P1: Add tool_result type to validateSessionHasOutput() to prevent
false negatives for tool-only background tasks that would otherwise
timeout after 30 minutes despite having valid results.
P2: Add pendingByParent tracking to resume() and registerExternalTask()
to prevent premature 'ALL COMPLETE' notifications when mixing
launched and resumed tasks.
* fix: address code review feedback - log messages, model passthrough, sorting, race condition
- Fix misleading log messages: 'promptAsync' -> 'prompt (fire-and-forget)'
- Restore model passthrough in launch() for Sisyphus category configs
- Fix call-omo-agent sorting: use time.created number instead of String(time)
- Fix race condition: check promptError inside polling loop, not just after 100ms
