451 Commits

Author SHA1 Message Date
YeonGyu-Kim
ae4e113c7e
Merge pull request #1610 from code-yeongyu/fix/96-compaction-dedup-recovery
fix: wire deduplication into compaction recovery for prompt-too-long errors (#96)
2026-02-07 19:28:49 +09:00
YeonGyu-Kim
403457f9e4 fix: rewrite dedup recovery test to mock module instead of filesystem 2026-02-07 19:26:06 +09:00
YeonGyu-Kim
5e5c091356
Merge pull request #1611 from code-yeongyu/fix/1481-1483-compaction
fix: prevent compaction from inserting arbitrary constraints and preserve todo state (#1481, #1483)
2026-02-07 19:23:50 +09:00
YeonGyu-Kim
1df025ad44 fix: use lazy storage dir resolution to fix CI test flakiness 2026-02-07 19:23:24 +09:00
YeonGyu-Kim
844ac26e2a fix: wire deduplication into compaction recovery for prompt-too-long errors (#96)
Ultraworked with [Sisyphus](https://github.com/code-yeongyu/oh-my-opencode)

Co-authored-by: Sisyphus <clio-agent@sisyphuslabs.ai>
2026-02-07 19:18:12 +09:00
YeonGyu-Kim
2727f0f429 refactor: extract context window recovery hook
Ultraworked with [Sisyphus](https://github.com/code-yeongyu/oh-my-opencode)

Co-authored-by: Sisyphus <clio-agent@sisyphuslabs.ai>
2026-02-07 19:17:55 +09:00
YeonGyu-Kim
3947084cc5 fix: add compaction todo preserver hook
Ultraworked with [Sisyphus](https://github.com/code-yeongyu/oh-my-opencode)

Co-authored-by: Sisyphus <clio-agent@sisyphuslabs.ai>
2026-02-07 19:15:46 +09:00
YeonGyu-Kim
67f701cd9e fix: avoid invented compaction constraints
Ultraworked with [Sisyphus](https://github.com/code-yeongyu/oh-my-opencode)

Co-authored-by: Sisyphus <clio-agent@sisyphuslabs.ai>
2026-02-07 19:15:41 +09:00
YeonGyu-Kim
6e2f3b1f50
Merge pull request #1593 from code-yeongyu/fix/prometheus-plan-overwrite
fix: allow Prometheus to overwrite .sisyphus/*.md plan files
2026-02-07 19:04:47 +09:00
YeonGyu-Kim
9a8f03462f fix: normalize resolvedPath before startsWith check
Addresses cubic review feedback — resolvedPath may contain
non-canonical segments when filePath is absolute, causing
the startsWith check against sisyphusRoot to fail.
2026-02-07 19:01:28 +09:00
YeonGyu-Kim
f980e256dd fix: boulder continuation now respects /stop-continuation guard
Add isContinuationStopped check to atlas hook's session.idle handler
so boulder continuation stops when user runs /stop-continuation.

Previously, todo continuation and session recovery checked the guard,
but boulder continuation did not — causing work to resume after stop.

Fixes #1575
2026-02-07 18:50:13 +09:00
YeonGyu-Kim
38169523c4 fix: anchor .sisyphus path check to ctx.directory to prevent false positives
- Uses path.join(ctx.directory, '.sisyphus') + sep as prefix instead of loose .includes()
- Prevents false positive when .sisyphus exists in parent directories outside project root
- Adds test for the false positive case (cubic review feedback)
2026-02-07 18:49:16 +09:00
YeonGyu-Kim
eb5cc873ea fix: trim whitespace from tool names to prevent invalid tool calls
Some models (e.g. kimi-k2.5) return tool names with leading spaces
like ' delegate_task', causing tool matching to fail.

Add .trim() in transformToolName() and defensive trim in claude-code-hooks.

Fixes #1568
2026-02-07 13:12:47 +09:00
YeonGyu-Kim
847d994199 fix: allow Prometheus to overwrite .sisyphus/*.md plan files
Add exception in write-existing-file-guard for .sisyphus/*.md files
so Prometheus can rewrite plan files without being blocked by the guard.

The prometheus-md-only hook (which runs later) still validates that only
Prometheus can write to these paths, preserving security.

Fixes #1576
2026-02-07 13:12:44 +09:00
YeonGyu-Kim
cb2169f334 fix: guard against undefined modelID in anthropic-effort hook
Add early return when model.modelID or model.providerID is nullish,
preventing TypeError at runtime when chat.params receives incomplete
model data.
2026-02-06 21:55:13 +09:00
YeonGyu-Kim
6febebc166 feat: add anthropic-effort hook to inject effort=max for Opus 4.6
Injects `output_config: { effort: "max" }` via AI SDK's providerOptions
when all conditions are met:
- variant is "max" (sisyphus, prometheus, metis, oracle, unspecified-high, ultrawork)
- model matches claude-opus-4[-.]6 pattern
- provider is anthropic, opencode, or github-copilot (with claude model)

Respects existing effort value if already set. Normalizes model IDs
with dots to hyphens for consistent matching.
2026-02-06 21:47:10 +09:00
YeonGyu-Kim
a691a3ac0a refactor: migrate delegate_task to task tool with metadata fixes
- Rename delegate_task tool to task across codebase (100 files)
- Update model references: claude-opus-4-6 → 4-5, gpt-5.3-codex → 5.2-codex
- Add tool-metadata-store to restore metadata overwritten by fromPlugin()
- Add session ID polling for BackgroundManager task sessions
- Await async ctx.metadata() calls in tool executors
- Add ses_ prefix guard to getMessageDir for performance
- Harden BackgroundManager with idle deferral and error handling
- Fix duplicate task key in sisyphus-junior test object literals
- Fix unawaited showOutputToUser in ast_grep_replace
- Fix background=true → run_in_background=true in ultrawork prompt
- Fix duplicate task/task references in docs and comments
2026-02-06 21:35:30 +09:00
YeonGyu-Kim
4692809b42 Regenerate AGENTS.md hierarchy with latest codebase state 2026-02-06 19:07:12 +09:00
YeonGyu-Kim
d8b29da15f fix(category-skill-reminder): dynamically include available skills with user priority 2026-02-06 19:03:06 +09:00
YeonGyu-Kim
60bbeb7304 fix(compaction): remove hardcoded Claude model from compaction hooks 2026-02-06 18:58:48 +09:00
YeonGyu-Kim
e9a3d579b3
Merge pull request #1553 from code-yeongyu/fix/1355-atlas-continuation-guard
fix(atlas): stop continuation retry loop on repeated prompt failures
2026-02-06 18:57:32 +09:00
YeonGyu-Kim
aec5624122 fix(atlas): stop continuation retry loop on repeated prompt failures 2026-02-06 17:34:14 +09:00
YeonGyu-Kim
ca8ec494a3 docs: fix stale references in AGENTS.md files 2026-02-06 17:20:19 +09:00
YeonGyu-Kim
25e436a4aa fix: update snapshots and remove duplicate key in switcher for model version update 2026-02-06 15:12:41 +09:00
YeonGyu-Kim
1f64920453 chore: update claude-opus-4-5 references to claude-opus-4-6 (excludes antigravity models) 2026-02-06 15:09:07 +09:00
YeonGyu-Kim
11d0005eb5 feat: prioritize claude-opus-4-6 over claude-opus-4-5 in anthropic fallback chains
Add claude-opus-4-6 as the first anthropic provider entry before
claude-opus-4-5 across all agent and category fallback chains.
Also add high variant mapping for think-mode switcher.
2026-02-06 03:31:55 +09:00
YeonGyu-Kim
6cfaac97b2
Merge pull request #1477 from kaizen403/fix/boulder-agent-tracking
fix: track agent in boulder state to fix session continuation (fixes #927)
2026-02-05 17:41:05 +09:00
YeonGyu-Kim
9800d1ecb0
Merge pull request #1424 from code-yeongyu/fix/auto-update-wrong-directory
fix(auto-update): use USER_CONFIG_DIR instead of CACHE_DIR for plugin invalidation
2026-02-05 02:31:14 +09:00
YeonGyu-Kim
ddf878e53c feat(write-existing-file-guard): add hook to prevent write tool from overwriting existing files
Adds a PreToolUse hook that intercepts write operations and throws an error
if the target file already exists, guiding users to use the edit tool instead.

- Throws error: 'File already exists. Use edit tool instead.'
- Hook is enabled by default, can be disabled via disabled_hooks
- Includes comprehensive test suite with BDD-style comments
2026-02-05 01:58:14 +09:00
YeonGyu-Kim
8886879bd0 fix(auto-update): use USER_CONFIG_DIR instead of CACHE_DIR for plugin invalidation
The auto-update-checker was operating on the wrong directory:
- CACHE_DIR (~/.cache/opencode) was used for node_modules, package.json, and bun.lock
- But plugins are installed in USER_CONFIG_DIR (~/.config/opencode)

This caused auto-updates to fail silently:
1. Update detected correctly (3.x.x -> 3.y.y)
2. invalidatePackage() tried to delete from ~/.cache/opencode (wrong!)
3. bun install ran but respected existing lockfile
4. Old version remained installed

Fix: Use USER_CONFIG_DIR consistently for all invalidation operations.

Also moves INSTALLED_PACKAGE_JSON constant to use USER_CONFIG_DIR for consistency.
2026-02-05 01:54:10 +09:00
Rishi Vhavle
38b40bca04 fix(prometheus-md-only): prioritize boulder state agent over message files
Root cause fix for issue #927:
- After /plan → /start-work → interruption, in-memory sessionAgentMap is cleared
- getAgentFromMessageFiles() returns 'prometheus' (oldest message from /plan)
- But boulder.json has agent: 'atlas' (set by /start-work)

Fix: Check boulder state agent BEFORE falling back to message files
Priority: in-memory → boulder state → message files

Test: 3 new tests covering the priority logic
2026-02-04 21:27:23 +05:30
Rishi Vhavle
169ccb6b05 fix: use boulder agent instead of hardcoded Atlas check for continuation
Address code review: continuation was blocked unless last agent was Atlas,
making the new agent parameter ineffective. Now the idle handler checks if
the last session agent matches boulderState.agent (defaults to 'atlas'),
allowing non-Atlas agents to resume when properly configured.

- Add getLastAgentFromSession helper for agent lookup
- Replace isCallerOrchestrator gate with boulder-agent-aware check
- Add test for non-Atlas agent continuation scenario
2026-02-04 21:21:57 +05:30
Rishi Vhavle
d8137c0c90 fix: track agent in boulder state to fix session continuation (fixes #927)
Add 'agent' field to BoulderState to track which agent (atlas) should
resume on session continuation. Previously, when user typed 'continue'
after interruption, Prometheus (planner) resumed instead of Sisyphus
(executor), causing all delegate_task calls to get READ-ONLY mode.

Changes:
- Add optional 'agent' field to BoulderState interface
- Update createBoulderState() to accept agent parameter
- Set agent='atlas' when /start-work creates boulder.json
- Use stored agent on boulder continuation (defaults to 'atlas')
- Add tests for new agent field functionality
2026-02-04 21:21:57 +05:30
YeonGyu-Kim
0dd42e2901 fix(non-interactive-env): force unix export syntax for bash env prefix
Ultraworked with [Sisyphus](https://github.com/code-yeongyu/oh-my-opencode)

Co-authored-by: Sisyphus <clio-agent@sisyphuslabs.ai>
2026-02-04 14:52:13 +09:00
YeonGyu-Kim
80e970cf36
Merge pull request #1297 from khduy/fix/deduplicate-settings-paths
fix(claude-code-hooks): deduplicate settings paths to prevent double hook execution
2026-02-04 13:35:06 +09:00
YeonGyu-Kim
d11f0685be
Merge pull request #1388 from boguan/dev
fix: remove redundant removeCodeBlocks call
2026-02-04 13:34:51 +09:00
Rishi Vhavle
6151d1cb5e
fix: block bash commands in Prometheus mode to respect permission config (#1449)
Fixes #1428 - Prometheus bash bypass security issue
2026-02-04 11:06:54 +09:00
YeonGyu-Kim
13e1d7cbd7
fix(non-interactive-env): use detectShellType() instead of hardcoded 'unix' (#1459)
The shellType was hardcoded to 'unix' which breaks on native Windows shells
(cmd.exe, PowerShell) when running without Git Bash or WSL.

This change uses the existing detectShellType() function to dynamically
determine the correct shell type, enabling proper env var syntax for all
supported shell environments.
2026-02-04 10:52:46 +09:00
YeonGyu-Kim
3a823eb2a2 feat(tasks-todowrite-disabler): add strong emphasis to register tasks before working
Add warning that even trivial tasks must be registered with TaskCreate
before starting work - no direct work without task tracking.
2026-02-03 16:27:58 +09:00
YeonGyu-Kim
4c4e1687da feat(auto-slash-command): add builtin commands support and improve part extraction
- Add builtin commands to command discovery with 'builtin' scope
- Improve extractPromptText to prioritize slash command parts
- Add findSlashCommandPartIndex helper for locating slash commands
- Add CommandExecuteBefore hook support
2026-02-03 14:33:53 +09:00
YeonGyu-Kim
1a0cc424b3 feat(tasks-todowrite-disabler): improve error message with actionable workflow guidance 2026-02-03 14:11:27 +09:00
YeonGyu-Kim
1b9303ba37
refactor(ultrawork): simplify workflow and apply parallel context gathering (#1412)
* refactor(ultrawork): simplify workflow to natural tool-like agent usage

Restore beta.16 style where explore/librarian agents feel like tools:
- Simplify delegate_task examples (agent=, background=true)
- Remove verbose DATA DEPENDENCIES explanation
- Condense EXECUTION RULES to action-oriented bullets
- Simplify WORKFLOW to 4 clear steps
- Remove procedural constraints that discouraged parallel exploration

The goal: agents fire background tasks AND continue direct exploration,
rather than waiting passively for background results.

* refactor(ultrawork/gpt5.2): apply two-track parallel context gathering

Based on GPT-5.2 Prompting Guide recommendations:
- 'Parallelize independent reads to reduce latency'
- Fire background agents (explore, librarian) for deep search
- Use direct tools (Grep, Read, LSP) simultaneously for quick wins
- Collect and merge ALL findings for comprehensive context

Pattern: background fire → direct exploration in parallel → collect → proceed

* fix: address Cubic review feedback

- Fix delegate_task parameter names in default.ts (agent → subagent_type, background → run_in_background)
- Add missing load_skills and run_in_background parameters to delegate_task examples
- Restore new_task_system_enabled property to schema and TypeScript config
- Fix tool names in gpt5.2.ts (Grep → grep, Read → read_file)

Identified by cubic (https://cubic.dev)
2026-02-03 12:13:22 +09:00
YeonGyu-Kim
ec1cb5db05
fix(prometheus): enforce path constraints and atomic write protocol (#1414)
* fix(prometheus): enforce path constraints and atomic write protocol

- Add FORBIDDEN PATHS section blocking docs/, plan/, plans/ directories
- Add SINGLE ATOMIC WRITE protocol to prevent content loss from multiple writes
- Simplify PROMETHEUS_AGENTS array to single PROMETHEUS_AGENT string

* fix: reconcile Edit tool signature in interview-mode.ts with identity-constraints.ts

Identified by cubic: Edit tool usage was inconsistent between files.

- interview-mode.ts showed: Edit(path, content)
- identity-constraints.ts showed: Edit(path, oldString="...", newString="...")

Updated interview-mode.ts to use the correct Edit signature with oldString and newString parameters to match the actual tool API and prevent agent hallucination.
2026-02-03 12:11:52 +09:00
YeonGyu-Kim
dea13a37a6
feat(task-system): add experimental task system with Claude Code spec alignment (#1415)
* feat(hooks): add tasks-todowrite-disabler hook to block TodoRead/TodoWrite

Ultraworked with [Sisyphus](https://github.com/code-yeongyu/oh-my-opencode)

Co-authored-by: Sisyphus <clio-agent@sisyphuslabs.ai>

* feat(task-tools): add parallel execution guidance to descriptions

Ultraworked with [Sisyphus](https://github.com/code-yeongyu/oh-my-opencode)

Co-authored-by: Sisyphus <clio-agent@sisyphuslabs.ai>

* refactor(index): migrate task system to experimental.task_system flag

Ultraworked with [Sisyphus](https://github.com/code-yeongyu/oh-my-opencode)

Co-authored-by: Sisyphus <clio-agent@sisyphuslabs.ai>

* docs: update AGENTS.md for experimental task system

Ultraworked with [Sisyphus](https://github.com/code-yeongyu/oh-my-opencode)

Co-authored-by: Sisyphus <clio-agent@sisyphuslabs.ai>

* fix(task-tests): align test field names with Claude Code spec (subject, blockedBy, addBlockedBy)

* fix: address Cubic review feedback

Ultraworked with [Sisyphus](https://github.com/code-yeongyu/oh-my-opencode)

Co-authored-by: Sisyphus <clio-agent@sisyphuslabs.ai>

* fix: add optional chaining for tasksTodowriteDisabler null check

---------

Co-authored-by: Sisyphus <clio-agent@sisyphuslabs.ai>
2026-02-03 12:11:23 +09:00
YeonGyu-Kim
ac9e22cce5 fix(prompts): add missing run_in_background and load_skills params to examples
All delegate_task examples now include required parameters to prevent
model confusion about parameter omission.

Fixes #1403
2026-02-03 10:50:26 +09:00
YeonGyu-Kim
7226836472 atlas reminder reinforce 2026-02-03 10:31:34 +09:00
YeonGyu-Kim
62e1687474 feat: add agent fallback and preemptive-compaction restoration
- Add agent visibility fallback for first-run scenarios
- Restore preemptive-compaction hook
- Update migration and schema for preemptive-compaction restoration
2026-02-02 22:40:59 +09:00
BoGuan
f68a6f7d1b
fix: remove redundant removeCodeBlocks call
Remove duplicate removeCodeBlocks() call in keyword-detector/index.ts.

The detectKeywordsWithType() function already calls removeCodeBlocks() internally, so calling it before passing the text was redundant and caused unnecessary double processing.
2026-02-02 15:18:25 +08:00
YeonGyu-Kim
134dc7687e fix(task-tool): add task ID validation and improve lock acquisition safety
- Add task ID pattern validation (T-[A-Za-z0-9-]+) to prevent path traversal
- Refactor lock mechanism to use UUID-based IDs for reliable ownership tracking
- Implement atomic lock creation with stale lock detection and cleanup
- Add lock acquisition checks in create/update/delete handlers
- Expand task-reminder hook to track split tool names and clean up on session deletion
- Add comprehensive test coverage for validation and lock handling
2026-02-01 23:50:34 +09:00
YeonGyu-Kim
8d29a1c5c7
Implement unified Claude Tasks system with single multi-action tool (#1356)
* chore: pin bun-types to 1.3.6

🤖 Generated with [OhMyOpenCode](https://github.com/code-yeongyu/oh-my-opencode)

* chore: exclude test files and script from tsconfig

🤖 Generated with [OhMyOpenCode](https://github.com/code-yeongyu/oh-my-opencode)

* refactor: remove sisyphus-swarm feature

Remove mailbox types and swarm config schema. Update docs.

🤖 Generated with [OhMyOpenCode](https://github.com/code-yeongyu/oh-my-opencode)

* refactor: remove legacy sisyphus-tasks feature

Remove old storage and types implementation, replaced by claude-tasks.

🤖 Generated with [OhMyOpenCode](https://github.com/code-yeongyu/oh-my-opencode)

* feat(claude-tasks): add task schema and storage utilities

- Task schema with Zod validation (pending, in_progress, completed, deleted)
- Storage utilities: getTaskDir, readJsonSafe, writeJsonAtomic, acquireLock
- Atomic writes with temp file + rename
- File-based locking with 30s stale threshold

🤖 Generated with [OhMyOpenCode](https://github.com/code-yeongyu/oh-my-opencode)

* feat(tools/task): add task object schemas

Add Zod schemas for task CRUD operations input validation.

🤖 Generated with [OhMyOpenCode](https://github.com/code-yeongyu/oh-my-opencode)

* feat(tools): add TaskCreate tool

Create new tasks with sequential ID generation and lock-based concurrency.

🤖 Generated with [OhMyOpenCode](https://github.com/code-yeongyu/oh-my-opencode)

* feat(tools): add TaskGet tool

Retrieve task by ID with null-safe handling.

🤖 Generated with [OhMyOpenCode](https://github.com/code-yeongyu/oh-my-opencode)

* feat(tools): add TaskUpdate tool with claim validation

Update tasks with status transitions and owner claim validation.

🤖 Generated with [OhMyOpenCode](https://github.com/code-yeongyu/oh-my-opencode)

* feat(tools): add TaskList tool and exports

- TaskList for summary view of all tasks
- Export all claude-tasks tool factories from index

🤖 Generated with [OhMyOpenCode](https://github.com/code-yeongyu/oh-my-opencode)

* feat(hooks): add task-reminder hook

Remind agents to use task tools after 10 turns without task operations.

🤖 Generated with [OhMyOpenCode](https://github.com/code-yeongyu/oh-my-opencode)

* feat(config): add disabled_tools setting and tasks-todowrite-disabler hook

- Add disabled_tools config option to disable specific tools by name
- Register tasks-todowrite-disabler hook name in schema

🤖 Generated with [OhMyOpenCode](https://github.com/code-yeongyu/oh-my-opencode)

* feat(config-handler): add task_* and teammate tool permissions

Grant task_* and teammate permissions to atlas, sisyphus, prometheus, and sisyphus-junior agents.

🤖 Generated with [OhMyOpenCode](https://github.com/code-yeongyu/oh-my-opencode)

* feat(delegate-task): add execute option for task execution

Add optional execute field with task_id and task_dir for task-based delegation.

🤖 Generated with [OhMyOpenCode](https://github.com/code-yeongyu/oh-my-opencode)

* fix(truncator): add type guard for non-string outputs

Prevent crashes when output is not a string by adding typeof checks.

🤖 Generated with [OhMyOpenCode](https://github.com/code-yeongyu/oh-my-opencode)

* chore: export config types and update task-resume-info

- Export SisyphusConfig and SisyphusTasksConfig types
- Add task_tool to TARGET_TOOLS list

🤖 Generated with [OhMyOpenCode](https://github.com/code-yeongyu/oh-my-opencode)

* refactor(storage): remove team namespace, use flat task directory

* feat(task): implement unified task tool with all 5 actions

* fix(hooks): update task-reminder to track unified task tool

* refactor(tools): register unified task tool, remove 4 separate tools

* chore(cleanup): remove old 4-tool task implementation

* refactor(config): use new_task_system_enabled as top-level flag

- Add new_task_system_enabled to OhMyOpenCodeConfigSchema
- Remove enabled from SisyphusTasksConfigSchema (keep storage_path, claude_code_compat)
- Update index.ts to gate on new_task_system_enabled
- Update plugin-config.ts default for config initialization
- Update test configs in task.test.ts and storage.test.ts

Ultraworked with [Sisyphus](https://github.com/code-yeongyu/oh-my-opencode)

Co-authored-by: Sisyphus <clio-agent@sisyphuslabs.ai>

* fix: resolve typecheck and test failures

- Add explicit ToolDefinition return type to createTask function
- Fix planDemoteConfig to use 'subagent' mode instead of 'all'

---------

Co-authored-by: justsisyphus <justsisyphus@users.noreply.github.com>
Co-authored-by: Sisyphus <clio-agent@sisyphuslabs.ai>
2026-02-01 22:42:28 +09:00