- Removed the `downshift.boomerang` feature, including CLI flags, configuration schema settings, and internal session state logic.
- Deleted the associated prompt definition file and all unit tests related to the boomerang validation flow.
- Cleaned up frontend forms and server request handling in the harbor-manager package to reflect the feature removal.
- Updated the downshift planning prompt instructions to maintain continuity in task creation.
- Modify downshift logic to ignore `todo` tool calls as triggers, requiring them instead to open a "gate" that permits switching only on subsequent `edit`/`write` actions.
- Ensure the starting model consistently handles implementation until the todo list is established, preventing premature hand-off to the fast/cheap model.
- Update system prompts to emphasize strict validation and multi-test execution requirements for the boomerang model upon returning to the primary context.
- Introduced `--downshift-boomerang` CLI flag and `downshift.boomerang` configuration setting.
- Integrated boomerang validation hand-back logic and context cutting within the agent session.
- Added system prompt templates and runtime checks to facilitate message handling for boomerang transitions.
- Included unit tests to verify context management and validation pass requirements during the boomerang process.
- Included the todo tool in downshift action triggers, gated on the plan nudge being in context — a post-nudge todo init is the planning-complete signal, while a turn-one todo remains bookkeeping.
- Updated flag help, settings schema, SDK docs, slash-command text, and plan-nudge instructions.
- Added a regression test covering the nudge-gated todo trigger.
- Replaced legacy reasoning-slide functionality with new downshift and plan-yolo capabilities.
- Updated CLI arguments, slash commands, and configuration schemas to support the new model-switching and execution behaviors.
- Refactored agent session logic to handle downshift arming, plan-yolo headless execution, and context scrubbing.
- Renamed and added system prompts to align with the updated downshift and plan-yolo workflows.
- Added a CLI usage-error type for argument validation failures.
- Reported invalid --max-time values from launch and ACP as clean usage errors with exit code 2.
- Covered the no-stack-trace CLI error path for invalid max-time input.
Fixes#5041
- Parsed --max-time values with s, m, and h suffixes into seconds instead of dropping the deadline.
- Raised a visible parse error for invalid or non-positive max-time values.
- Updated help text and regression coverage for the CLI timeout parser.
Fixes#5041
- Added `off` and `auto` as valid inputs for the `--thinking` CLI flag.
- Centralized thinking level definitions in `CLI_THINKING_LEVELS` to keep flag options, shell completions, and validation in sync.
- Configured CLI parsing to reject `inherit` as an explicit input to prevent unintended configuration suppression.
- Added a new `--advisor` flag to argument parsing and launch flag definitions.
- Applied the parsed `advisor` flag as a runtime-only override of `advisor.enabled` during startup.
- Updated advisor-related docs and added parseArgs tests for `--advisor` behavior and position handling.
- Added `omp models` command with `ls`, `find`, `canonical`, and `refresh` actions.
- Removed top-level `--list-models` parsing from CLI args, launch, and main command flow.
- Implemented action-driven model listing with provider filtering, extension loading, and `--json` output.
- Updated unknown provider/model errors and tests to direct users to `omp models` guidance.
Upstream force-rewrote history; this branch carried old-SHA twins of the
rewritten commits. All non-goal conflicts resolved to upstream (verified
ours == old upstream tip). Goal-side reconciliation:
- cli.ts: profile bootstrap woven into the new lazy-import/resolveCliArgv
structure; worker-host entry declaration deferred until after profile
selection (pi-utils/env eagerly snapshots the agent dir .env); the
floating runCli call guarded with import.meta.main || !Bun.isMainThread
so importing runCli stays side-effect free while Worker re-entry works.
- args.ts/flag-tables.ts: kept profile/alias branches; upstream's new
repeatable --config overlay flag moved into STRING_SETTERS.
- Changelogs: upstream-released bullets deduped out of Unreleased; profile
entries restored under Unreleased.
- task/index.ts: removed duplicated validateTaskIds block from auto-merge.
Move bundled models, model cache/manager, thinking metadata, effort helpers,
provider descriptors/discovery, wire constants, and model identity utilities
into the new @oh-my-pi/pi-catalog package.
Update pi-ai to keep provider runtime/auth concerns, move catalog provider
metadata into CATALOG_PROVIDERS, and migrate coding-agent, agent, stats, docs,
and tests to import catalog values from pi-catalog.
Split coding-agent model registry helpers into discovery, roles, and models
config modules while preserving registry orchestration.
BREAKING CHANGE: @oh-my-pi/pi-ai no longer exports catalog subpaths such as
/models, /model-cache, /model-manager, /model-thinking, /effort,
/provider-models*, discovery helpers, and provider wire constants; use the
matching @oh-my-pi/pi-catalog subpaths instead.
- Migrated Effort and THINKING_EFFORTS imports to @oh-my-pi/pi-ai/effort in CLI args and launch command files.
- Split model-registry dependencies across focused @oh-my-pi/pi-ai submodules instead of the root barrel export.
Added named OMP profiles that isolate agent state (auth credentials,
sessions, settings, model cache, history, memories, blobs, plus
config root subdirs) under `~/.omp/profiles/<name>/agent/`. Activated
via `--profile <name>` or `OMP_PROFILE=<name>`; `default` maps back to
the regular `~/.omp/agent/` tree.
Added `--alias <command>` to generate a shell shortcut (e.g.
`omp-work`) that forwards `omp --profile <name>`. Detects the active
shell (bash, zsh, fish, PowerShell, pwsh), writes a wrapper into the
correct rc file, and preserves subcommands like `update`, `--version`,
and `--model` because the wrapper passes through argv unchanged.
The `--profile`/`--alias` bootstrap pre-parser lives in
`packages/coding-agent/src/cli/profile-bootstrap.ts` and runs before
any module that touches `getAgentDir()` (notably `@oh-my-pi/pi-utils/env`,
which eagerly loads `.env` from the agent directory at its own import
time). The pre-parser mirrors `parseArgs` value-consumption rules and
honors `--`, so commands like `omp --system-prompt --profile foo` pass
the literal `--profile` through as the prompt body instead of silently
activating profile `foo`.
XDG resolution for named profiles is keyed on the profile-specific
XDG path (`$XDG_*_HOME/omp/profiles/<name>`), never the base app root,
so a profile's location is decided once at first activation and stays
stable even after `omp config init-xdg` materializes the base later.
The default profile keeps its existing base-app-root check.
`setProfile(undefined)` (and `setProfile("default")`) restores the
pre-profile `PI_CODING_AGENT_DIR` snapshot taken at first activation
instead of unconditionally deleting it. `setAgentDir` refreshes the
snapshot since that call is the user explicitly redefining the
baseline.
Validation rejects profile names that match `.`/`..`, fail
`/^[A-Za-z0-9][A-Za-z0-9._-]{0,63}$/`, or hit a Windows reserved
device name (`CON`, `PRN`, `AUX`, `NUL`, `COM0-9`, `LPT0-9`, including
dotted variants like `CON.txt`) — those would let `setProfile` accept
the input only for directory creation to fail later with confusing
errors on Windows.
- Added `ToolTier`, `ToolApproval`, and `ToolApprovalDecision` types and exported approval APIs.
- Updated approval-mode options from `auto|prompt|custom` to `always-ask|write|yolo` and defaulted mode to `yolo`.
- Changed approval resolution to apply per-tool decisions first, then mode-tier limits, with legacy-mode migration.
- Assigned read/write/exec `approval` and approval-detail prompts across built-in, custom, extension, and MCP tools.
- Added a new `approvalMode` argument to CLI parsing with validation for `auto`, `prompt`, and `custom` values.
- Registered `--approval-mode` on the launch command so it appears in generated help output.
- Applied the parsed approval mode as a runtime override on `Settings`, ensuring downstream `tools.approvalMode` reads reflect the CLI value.
- Decouple the per-tool approval gate from extension presence. ExtensionRunner
and the ExtensionToolWrapper that hosts the gate are now constructed
unconditionally in createAgentSession. Previously the runner was only built
when extensionsResult.extensions.length > 0, so the entire approval system
silently disappeared for sessions with no extensions loaded — any
tools.approvalMode: prompt|custom setting was a no-op without feedback.
Today this hole was masked by createAutoresearchExtension always being
pushed inline; the unconditional construction makes the safety invariant
explicit, and a new regression test in approval-mode.test.ts pins it.
- Extend CRITICAL_BASH_PATTERNS to cover remote-fetch-then-execute shapes
that the original `bash <(curl …)` regex missed:
- `source <(curl …)` / `. <(curl …)` (anchored at command boundary so
`find . -name foo` doesn't false-positive)
- `eval "$(curl …)"` / `eval $(curl …)` / `eval `curl …``
Also adds `chmod -R` symbolic-mode forms (`u+x`, `u+rwx,o+w …`) targeting
filesystem root, and `tee` / `tee -a` writes to /etc/{passwd,shadow,sudoers}
(the standard way to write root-owned files without redirect). Benign
forms (`source ./local.sh`, `chmod -R u+x ./build`, `tee /var/log/app.log`,
`eval "$VAR"`) are pinned negative in the test suite.
- Extend formatApprovalPrompt with payload previews for the destructive tools
that previously rendered as bare `Allow tool: <name>`: eval (language +
first cell's code), task (agent + first task's id + assignment), ast_edit
(first op's pattern / replacement / paths), browser (action + tab + url +
code), and write content (alongside path). For `task` in particular this
closes the gap that docs/approval-mode.md's "parent's approval covers the
subagent" claim was waving at — the prompt now actually shows what's being
delegated.
- Tighten isMcpToolName: drop the fallback `|| toolName.includes("__")` so
an extension tool legally named `my__feature` or `pkg__util__do` is no
longer falsely labelled `Origin: MCP server tool` in the approval prompt.
Strict `mcp__` prefix only.
- Revert the cargo-cult `{ autoApprove: true } as AgentToolContext` insertions
in agent-session-python-cleanup.test.ts and sdk-move-cwd.test.ts. The tests
create sessions without passing settings, so the wrapper falls through to
approvalMode "auto" automatically; the explicit flag was unnecessary and
the `as AgentToolContext` cast hid that autoApprove lives on
CustomToolContext, not AgentToolContext.
- Document in commands/launch.ts the dual --auto-approve declaration (oclif
Flags for --help, manual parseArgs for runtime) so a future rename catches
both call sites.
- Promote the subagent caveat in docs/approval-mode.md to a callout near the
top: anything `task` is asked to do runs unattended once the parent task
call is approved.
Verification:
- bun test packages/coding-agent/test/tools/approval.test.ts → 75 pass / 0 fail
(was 57; +18 cases covering new remote-exec patterns, chmod symbolic, tee
/etc, isMcp negative, and eval/task/ast_edit/browser/write payload previews)
- bun test packages/coding-agent/test/tools/approval-mode.test.ts → 7 pass /
0 fail (was 7; +1 case asserting extensionRunner is always constructed)
- bun tsc --noEmit -p packages/coding-agent → clean
- bun x biome check . → clean
- Windows EBUSY tempdir-cleanup noise in agent-session-python-cleanup and
sdk-move-cwd is pre-existing on this branch (already documented in the
PR body) and absent on Linux CI.
Re-introduces the per-tool approval system from luzidd's commit 39124f3 (which
is no longer reachable from main) and improves it before re-landing.
What's restored:
- ApprovalPolicy (allow/deny/prompt) plus DEFAULT_APPROVAL_POLICIES.
- ACTION_EXCEPTIONS registry (LSP read-only, bash critical patterns).
- getApprovalPolicy() six-level resolution order.
- ExtensionToolWrapper.execute() gate before extension handlers.
- --auto-approve / --yolo CLI flag and tools.approval.<tool> user config.
- docs/approval-mode.md user guide.
What's improved over the original:
- Replaced unchecked 'as any' casts with typed unknown narrowing helpers.
- Validate userConfig values: invalid strings, numbers, etc. fall through to
the built-in default instead of being silently honoured (typo no longer
locks a tool out or grants implicit approval).
- Expanded CRITICAL_BASH_PATTERNS: chmod -R /, chown -R /, bash <(curl ...),
writes to /etc/passwd|shadow|sudoers, shutdown/reboot/halt/init 0,
kill -9 1, nc -e / nc -c reverse shells. Pattern shapes require a
command-position boundary so 'npm run reboot-tests' and 'echo "shutdown the
queue"' don't false-positive.
- Added DEBUG_READONLY_ACTIONS exception so DAP inspection actions (threads,
stack_trace, variables, scopes, read_memory, …) auto-allow while
execution-side actions (launch, attach, continue, evaluate, write_memory,
set_breakpoint, …) still prompt.
- formatApprovalPrompt: labels mcp__<server>__<tool> calls as MCP server
tools, surfaces ssh host + command, recognises the modern § hashline header
for edit, and truncates >240-char fields so a heredoc-sized body cannot
blow out the confirmation dialog.
- Test suite grown from 40 to 57 cases — new coverage for invalid user
config, the extended critical-bash patterns, benign-keyword negatives,
debug exceptions, MCP/ssh prompt formatting, and command truncation.
Verification:
- bun test packages/coding-agent/test/tools/approval.test.ts -> 57 pass
- bun x biome check . -> clean
- bun run check:ts across all 9 workspaces -> clean
Wire --hide-thinking launch flag that sets hideThinkingBlock before TUI
init. Display-only: does not disable model reasoning, just hides the
thinking output in the terminal.
- Add hideThinking to Args interface and parseArgs
- Add --hide-thinking flag definition in launch command
- Apply setting via settingsInstance.override in main
Closes#1313
- Added `omp auth-gateway serve/token/status` — a forward-proxy injecting broker credentials for OpenAI Chat, Anthropic Messages, and OpenAI Responses wire formats.
- Added `GET /v1/usage` to auth-broker and auth-gateway; usage cache switched to 5-min per-credential TTL with jitter and last-good fallback on failure.
- Added `AuthStorage.setConfigApiKey/removeConfigApiKey/clearConfigApiKeys` so `models.yml` `apiKey` beats OAuth tokens without overriding `--api-key`.
- Added `omp auth-broker migrate --from-local` for idempotent upload of local SQLite/env credentials to the broker.
- Adds omp acp subcommand that launches the agent as an ACP stdio server
- Registers the subcommand in the CLI dispatcher
- Threads terminal-auth args and ACP flags through the launch and main orchestrators
- Exports AgentSession on the public SDK surface
- Updates skills loader to support skill→slash-command conversion and prompt injection
- Updates input-controller to dispatch ACP built-in slash commands
Adds a new `rpc-ui` mode that extends the existing headless RPC mode with
interactive tool support (ask tool, extension UI dialogs, etc.).
In plain `rpc` mode the session has `hasUI=false` and no UI context is
wired, so interactive tools are disabled. `rpc-ui` mode sets `hasUI=true`
and wires a single shared `RpcExtensionUIContext` instance into both the
tool context store and the extension runner. Both consumers share the same
`pendingExtensionRequests` map and output closure, so `extension_ui_response`
messages received on stdin are routed to the correct waiting promise
regardless of which code path (tool or extension) created the request.
Changes:
- `args.ts`: add `rpc-ui` to the `Mode` union and the parse guard
- `launch.ts`: expose `rpc-ui` in the OCLIF flag definition and help text
- `main.ts`: propagate `rpc-ui` through all RPC-mode guard conditions and
pass `setToolUIContext` to `runRpcMode` when the mode is `rpc-ui`
- `rpc-mode.ts`: accept optional `setToolUIContext` callback; create one
shared `RpcExtensionUIContext` instance and pass it to both the tool
context store and the extension runner
- Introduced Effort enum and ThinkingConfig metadata for per-model reasoning capabilities with min/max effort levels.
- Migrated thinking level API from string-based ThinkingLevel to structured Effort enum across agent and AI packages.
- Added model-thinking module with effort mapping, policy application, and semantic versioning utilities for provider-specific thinking modes.
- Removed supportsXhigh() function and replaced effort clamping with model-aware validation using ThinkingConfig metadata.
- Expanded models.json with thinking configuration objects for 50+ models including Claude, Gemini, and OpenAI variants.
- Added Python analysis scripts for edit tool usage patterns and tool invocation stream processing.
- Removed static thinking mode constant exports (THINKING_LEVELS, ALL_THINKING_LEVELS, ALL_THINKING_MODES, THINKING_MODE_DESCRIPTIONS, THINKING_MODE_LABELS) in favor of dynamic function-based API.
- Renamed formatThinking() to getThinkingMetadata() with return type changed from string to structured ThinkingMetadata object containing value, label, and description.
- Renamed getAvailableThinkingLevel() to getAvailableThinkingLevels() and getAvailableThinkingEffort() to getAvailableThinkingEfforts() with added default parameters for runtime flexibility.
- Updated all consumer modules to use new function-based API instead of static constants, enabling dynamic thinking mode configuration.
- Extracted thinking module with ThinkingEffort, ThinkingLevel, and ThinkingMode types to centralize reasoning configuration across packages.
- Migrated ThinkingLevel type from pi-agent-core to pi-ai package with new validation functions parseThinkingLevel() and getAvailableThinkingLevel().
- Consolidated thinking level constants and descriptions into reusable exports (ALL_THINKING_LEVELS, THINKING_MODE_DESCRIPTIONS) for consistent UI display.
- Removed local thinking-effort-label utility and replaced formatThinkingEffortLabel() with centralized formatThinking() function from pi-ai.
- Refactored thinking mode handling to distinguish ThinkingSelector (user-facing with 'off' option) from ThinkingEffort (provider-level).
- Consolidated @oh-my-pi/pi-utils subpath imports into single package root import across 100+ files.
- Moved tryParseJson utility from local web scrapers module to @oh-my-pi/pi-utils package for centralized JSON parsing.
- Renamed loadSkillsFromDir to scanSkillsFromDir and refactored skill discovery to use fs.promises.readdir instead of glob-based approach.
- Replaced custom parseJSON with tryParseJson across discovery modules for consistent error handling.
- Removed emitCustomToolSessionEvent method and cleanupSshResources function, consolidating shutdown logic into dispose method.
- Updated glob pattern construction to use GlobBuilder with literal_separator(true) for improved path handling.
- Added `--no-rules` CLI flag to coding-agent to disable rules discovery and loading.
- Added `rules` option to CreateAgentSessionOptions to allow custom rules configuration.
- Added `sessionDir` option to RpcClientOptions and implemented Symbol.dispose() for resource cleanup.
- Removed tarball-based task loading; migrated to directory-based fixtures with required inputDir and expectedDir properties.
- Refactored runner to use RpcClient resource management with `using` statement and simplified fixture handling.
- Consolidated type definitions and removed tarball.ts module in favor of streamlined task interface.
- Extracted directory path utilities from multiple packages into a centralized '@oh-my-pi/pi-utils/dirs' module.
- Moved 30+ path helper functions (getAgentDir, getConfigRootDir, getPluginsDir, getMCPConfigPath, etc.) from scattered locations into a single shared utility module.
- Consolidated APP_NAME, CONFIG_DIR_NAME, and VERSION constants into the centralized dirs module for reuse across packages.
- Updated 70+ import statements across packages/ai, packages/coding-agent, packages/stats, and packages/tui to use the new centralized module.
- Removed local path construction logic and replaced with utility function calls for improved maintainability and consistency.
- Deleted packages/coding-agent/src/extensibility/plugins/paths.ts as its functions were moved to the centralized dirs module.
- Added PTY (pseudo-terminal) backed interactive command execution with streaming output support via PtySession class in pi-natives.
- Added PtyStartOptions and PtyRunResult types to configure and report PTY session execution status.
- Added write(), resize(), and kill() methods to PtySession for interactive control of running commands.
- Added interactive bash execution via PTY with real-time terminal rendering and input forwarding in bash-interactive tool.
- Added --no-pty CLI flag and PI_NO_PTY environment variable to disable PTY-based interactive bash execution.
- Added bash.virtualTerminal setting to control PTY-backed interactive execution behavior.
- Created lightweight CLI framework in @oh-my-pi/pi-utils/cli as drop-in replacement for oclif with zero dependencies.
- Migrated coding-agent CLI from oclif framework to @oh-my-pi/pi-utils/cli with lazy command loading and simplified architecture.
- Changed default root command from 'index' to 'launch' for improved UX.
- Removed @oclif/core and @oclif/plugin-autocomplete dependencies from coding-agent package.
- Deleted custom oclif help renderer (oclif-help.ts) as help rendering is now integrated into the new CLI framework.
- Moved launch command from nested index/index.ts to top-level launch.ts for clearer command structure.
can1357