- Added `AgentRegistry` singleton with session registration/unregistration and IRC routing metadata for peer lookups.
- Added IRC messaging prompts and tooling with `irc.enabled` setting, `list/send` tool paths, and peer roster rendering.
- Changed `/btw` to session-side `runEphemeralTurn`, added background IRC exchange flushing, and fixed empty-input checks.
- Added unit tests for IRC tool and BtwController ephemeral behavior, including disabled, busy, not-found, and abort cases.
- Removed `pi-natives` chunk language classifier modules and all core chunk subsystems (kind, state, render, edit, resolve).
- Removed chunk-mode CLI/read/edit entrypoints, including `read` command and chunk mode registration/prompt tooling.
- Removed chunk selectors from `read` and `grep` tools, switching behavior to raw/L-range handling.
- Fixed poll wait parsing to keep defaulting to `30s` when the provided value is empty.
- Added a new `async.pollWaitDuration` setting with enum values for 5s to 5m and defaulted it to 30s.
- Updated PollTool to parse the configured wait duration and include a timeout in the job-promise race before returning.
- Revised the poll tool prompt to describe timeout return behavior and discourage indefinite unproductive polling.
- Added `sed` atom editing with `g`, `i`, and `F` flags, path/anchor parsing, and conflict handling updates.
- Changed hashline and grep/read output to `LINE+ID|content` with `>` match prefixes and `:` context prefixes.
- Fixed atom anchor parsing for path-qualified locs, hyphenated single anchors, and content hints after `|` or `:`.
- Updated prompts, changelog, and tests to document and validate the new hashline and `sed` formats.
- Consolidated all former gh_* tools into a single GithubTool that routes execution by a required op field.
- Replaced gh_* tool registrations and render dispatch with `github`, including renderer key and header updates.
- Removed deprecated gh-* tool prompt files and added a unified github.md prompt covering per-op inputs and outputs.
- Updated settings-schema and ci-green prompt guidance to reference the unified github tool and `run_watch` operation.
- Updated tests and tool imports to use `GithubTool`/`githubToolRenderer` with op-based payloads and assertions.
- Added `HASHLINE_CONTENT_SEPARATOR` and switched hashline formatting to colon separators across hashline helpers.
- Updated hashline prefix regexes in `edit/modes/hashline.ts` to require `:` and stop accepting tab separators.
- Updated `read.ts` and `write.ts` hashline prepending/stripping to match `LINE+ID:` content prefixes.
- Updated `match-line-format.ts` and hashline read-mode docs to document and emit `LINE+ID:content` lines.
- Added `openai` and `openai-codex` as image providers and let `providers.image=auto` prefer GPT images.
- Updated settings, selector, and SDK wiring so OpenAI image providers pass through `setPreferredImageProvider`.
- Replaced Gemini-only image tooling with `image-gen` and added OpenAI/Codex hosted-image execution with SSE parsing.
- Added image-gen and handoff tests, including final-yield no-compaction regression and OpenAI payload/header assertions.
- Expanded hashline and chunk bigram tables to 647 entries and moved chunk checksums to a 40-item namespace.
- Changed hashline anchors from `LINE#ID`/`:` to concatenated `LINEID`\t forms across parsing and tool outputs.
- Removed line-number padding and routed diff/read/grep/renderer output through raw numbers, tabs, and `toDisplayLine` formatting.
- Renamed atom ops to `pre`/`post`, removed `ins`, and updated schemas, prompts, and tests for new insertion behavior.
- Canonicalized file and CLI defaults from `read` to `open` across tool registration and prompts.
- Added `resolveToolAlias()` and applied alias-normalized tool selection so legacy `read` maps to `open`.
- Updated runtime, UI, and export layers to treat `open` as first-class while preserving `read` compatibility.
- Renamed read prompt docs to `open.md`/`open-chunk.md` and refreshed system guidance to recommend `open`.
- Updated tool-related tests and expectations from `read` to `open` (including test fixtures and aliases).
- Added configurable minimizer settings to native and coding-agent APIs, including shellMinimizer options.
- Added minimizer execution plumbing through shell config, session-key generation, and buffered output capture.
- Added command identity detection and dispatching by program/subcommand with safe fallback to passthrough output.
- Added filter suites for git, docker, go, bun, cloud, and system commands to strip ANSI and compact noisy output.
- Added unit tests for detection and minimizer behavior across command support, failures, and passthrough paths.
Mirror the existing `commands.enableClaudeUser`/`commands.enableClaudeProject`
schema entries so the OpenCode discovery provider exposes the same
user/project toggle surface as Claude. Default remains true to preserve
current behavior.
Fixes#661
- Added built-in model entries for gpt-5.5 and gpt-image-2, including updated context windows, token limits, and pricing.
- Updated generated-model policy application to set or clear applyPatchToolType based on inferred GPT-5 freeform rules.
- Changed Spark edit-mode resolution to return apply_patch by default, honoring explicit replace and strict-mode overrides.
- Added tests for GPT-5 freeform policy inference and Spark edit-mode default, variant, and strict-mode behavior.
Slots a new "apply_patch" variant alongside the existing edit modes
(replace, patch, hashline, chunk, vim). The mode accepts a single input
string containing a Codex *** Begin Patch / *** End Patch envelope,
parses it with a new lenient parser (heredoc-tolerant), and fans each
file-op out to the existing executePatchSingle so LSP writethrough,
plan-mode guards, fs-cache invalidation and diagnostics are shared
with the patch mode.
Exposes both tool shapes from the spec: the JSON function-tool variant
(§1.2, {input: string}) and the OpenAI custom-tool / Lark-grammar
"freeform" variant (§1.1, raw patch string). The edit tool advertises
a Lark grammar via customFormat and a wire name via customWireName;
openai-responses emits it as a grammar-constrained custom tool when a
model opts in with applyPatchToolType: "freeform" in models.json.
custom_tool_call / custom_tool_call_output are plumbed end-to-end
through the shared responses code (emission, streaming, history
replay), and the agent-loop dispatcher matches tool calls by either
name or customWireName so returned calls route correctly.
Also threads preview/diff rendering for apply_patch through the TUI
(tool-execution + edit renderer) so streaming patches show per-file
diffs like the other edit modes.
Default edit mode is unchanged (hashline); opt in via edit.mode or
PI_EDIT_VARIANT=apply_patch.
Further compresses packages/coding-agent/src/prompts/system/system-prompt.md
from 21,530 B to 16,545 B (−23% on top of the prior compression;
−9,060 B / ~−2,265 tok per turn vs origin/main) while restoring
RFC 2119 weight on every inviolable rule and reorganizing the
pre-yield discipline.
system-prompt.md
RFC 2119 keyword restoration
All rules under `# Contract`, Procedure §§2/6, and the tail `<critical>`
block use `**MUST**` / `**MUST NOT**` keywords that the file's preamble
pins to RFC 2119. The previous draft had downgraded 29 of these to
`Do **NOT**`, which reads as below-MUST-NOT against the same preamble.
Restored sites:
- `# Contract` — all 7 inviolable bullets
- `<critical>` tail block — all 4 safety rules
- `## 6. Verification` — mock ban and no-proof yield rule
- Procedure §2 — "search for existing examples" rule regains the
PROHIBITED parallel-convention clause it lost
- `<dir-context>` — AGENTS.md read requirement regains MUST
Structural dedup
- `<source-of-truth>` and `<instruction-priority>` were two priority
lists with overlapping scope. Merged into a single
`<instruction-priority>` covering all 5 conflict-resolution levels.
- `<self-check>` and `<scope-check>` were two pre-yield checklists;
`## 6. Verification` also had a third numbered "Before yielding,
verify..." list with overlapping checks. Collapsed all three into
one `<pre-yield-check>` section (incl. "output format matches the
ask" from the old verification list). `## 6. Verification` now
focuses on verification discipline (test rigor, mocks, run-scope).
Signal restoration
Added four anchors to `<design-checklist>` that were dropped from the
earlier `<code-integrity>` block and had no home in the new structure:
- Adversarial-caller / tired-maintainer self-questioning
- Cost-of-easy-path framing
- Inhabit-the-call-site self-review
- Persistence on hard problems (do not punt half-solved work)
Handlebars fix
The `### Tool priority` header was emitted unconditionally but its
body was wrapped in `{{#ifAny python|bash}}`, producing a bare heading
with no content when neither tool was available. Header now sits
inside the conditional.
SECTION_SEPARATOR helper relocation
The `SECTION_SEPARATOR` Handlebars helper is a generic section-header
formatter that was registered in coding-agent's
`config/prompt-templates.ts`. This coupled every template consumer to
a side-effect import of that module; a prior fix added
`import "./config/prompt-templates"` to `system-prompt.ts` so the
`/system-prompt` sub-path export would register the helper, but
sibling consumers (`task/executor.ts`, `task/template.ts`) worked only
by accident because the parent agent happened to load
`system-prompt.ts` first.
- Moved `sectionSeparator` function and helper registration to
`packages/utils/src/prompt.ts` next to the other generic helpers
(`xml`, `codeblock`, `ifAny`, `includes`, `not`, `jsonStringify`).
- `packages/coding-agent/src/config/prompt-templates.ts` now
re-exports `sectionSeparator` from `@oh-my-pi/pi-utils/prompt` for
the test that imports it via the coding-agent path.
- Removed the side-effect import from
`packages/coding-agent/src/system-prompt.ts`.
Template typo fix
Also fixes a latent `SECTION_SEPERATOR` spelling (→ `SECTION_SEPARATOR`)
in the two subagent templates (`subagent-system-prompt.md`,
`subagent-user-prompt.md`). Before the move the typo was masked
because every call site and the helper shared the wrong spelling.
Post-move, the helper is canonical `SECTION_SEPARATOR` in pi-utils
and all templates match.
Verification
- `bun check` clean (TS + Rust, all 9 packages)
- `bun test packages/coding-agent/test/system-prompt-templates.test.ts`
6/6 pass (was 4/6 failing pre-existing before the infrastructure fix)
- `bun test packages/coding-agent/test/tools/task-template.test.ts`
4/4 pass (exercises the `sectionSeparator` re-export)
- `bun run format-prompts` clean
When a user models.yml declares a provider with auth: none (e.g. from
an older version), getApiKey() returns the literal 'N/A' without ever
consulting authStorage — even when valid OAuth credentials exist from
a successful /login.
Check authStorage.hasAuth() before returning kNoAuth. If real
credentials are present, fall through to the normal auth flow.
Fixes#749
- Added a new read.toolResultPreview boolean setting defaulting to false to control inline read result rendering.
- Updated read tool group rendering to show inline previews only when enabled and to hide duplicate summary rows for previewed entries.
- Passed the setting through event and UI helper constructors and added tests for default-off behavior and the duplicate-preview summary case.
- Added task.simple to settings and schema with default, schema-free, and independent modes.
- Added mode-aware task schema and validation to enforce context/schema rules per simple mode.
- Updated prompts and template rendering to tailor headers and guidance for each simple mode.
- Added simple-mode capabilities and updated TaskTool execution for mode-aware context behavior.
- Added tests for independent rendering and mode-specific rejection of invalid context or schema inputs.
Add SearXNG as a search backend for the web_search tool, enabling
privacy-friendly, self-hosted metasearch without API keys or usage limits.
Changes:
- New SearXNGProvider implementing SearchProvider interface
- Supports bearer token auth, categories, language, and recency filters
- Configurable via settings (searxng.endpoint, .token, .categories, .language)
or environment variables (SEARXNG_ENDPOINT, SEARXNG_TOKEN)
- Registered in provider chain and settings schema
- Maps SearXNG JSON API response to unified SearchResponse
Closes#545
- Added `toolStrictMode` support with `all_strict`/`none`/`mixed` options to OpenAI compatibility.
- Fixed OpenAI-completion strict-mode flows by capturing failed HTTP responses and retrying once as non-strict.
- Fixed completion error reporting by surfacing captured status, headers, and JSON `type`/`param`/`code` details.
- Improved strict-schema enforcement with WeakMap memoization and circular-schema detection in sanitization.
- Fixed OpenRouter provider lookup by resolving fallback model IDs for suffix and date variants in registry resolution.
- Refactored benchmark tooling and added async RPC error-window tracking for scheduled run execution.
- Added `vim` as a supported edit mode and resolved mode-specific active tool names.
- Migrated chunk-edit requests from legacy `op`/`content` to `path` plus `write`, `replace`, and `insert`.
- Replaced chunk-edit internals with `put`/`replace`/`delete` operations and ID-based selection.
- Updated chunk prompts/output docs to document ID-based selectors and `write: null` deletions.
- Switched chunk and grep render output to `@`/dash-style anchors and updated file-count behavior for multi-file edits.
- Updated multi-file result rendering with distinct file counting and pending-file progress indicator.
- Refreshed chunk-mode and chunk-tree tests for the new contract and anchor style.
- Added `/rename <title>` slash command to set explicit session names and update header/tab titles.
- Added `session_name` status segment with hash-derived accent color for session titles.
- Fixed shell execution failure sanitization to preserve all `execResult` fields after redacting stderr.
- Fixed tool execution completion output to pass original `toolResult` text instead of sanitized `content`.
- Add /rename <title> slash command to set an explicit session name,
updating the session header, terminal tab title, and status line
- Add session_name status segment: displays the session name on the
right side of the status bar with a stable djb2-hash-derived accent
color unique to each name; shown in all presets when a name is set
- Add setSessionName source tracking ('user' vs 'auto'): auto-generated
titles are silently ignored once the user has explicitly set a name,
preventing the async title generation from overwriting a /rename
- Sanitize session names at storage time: strip C0/C1 control characters
(including ANSI ESC) via static #sanitizeName before storing, blocking
escape sequence injection into the TUI, terminal title, and session file
- Extend sanitizeStatusText to cover the full C0/C1 range as
defense-in-depth (superset of the previous \r\n\t-only strip)
- Use stored (sanitized) name for showStatus message and terminal title
in handleRenameCommand, not the raw user input
- Guard terminal title update in auto-title path via titleSource === 'auto'
so a user rename is never overwritten by a late-resolving LLM call
- Reset #titleSource in #newSessionSync so each new session starts fresh
- Thread source parameter through AgentSession.setSessionName wrapper;
extension API and RPC set_session_name treated as 'user' intent
Closes#658
- Added bash.autoBackground.enabled and bash.autoBackground.thresholdMs settings with defaults for background job behavior.
- Added auto-backgrounding support for long foreground bash commands with managed job state and timeout threshold.
- Updated bash prompts, job-protocol messages, and tool activation checks to use async and auto-background support.
- Added background bash completion integration with AsyncJobManager and end-to-end tests for short/long auto-background scenarios.
- Added canonical model equivalence types, cache helpers, and registry APIs for provider variant lookup.
- Changed model resolution to apply canonical ID overrides/excludes with provider order before fallback matching.
- Added canonical and provider model views in list-models and selector UI with canonical sorting/persistence.
- Updated role/model persistence to store selectors while runtime now resolves concrete canonical-backed provider models.
Extension-registered models (via registerProvider) were stored directly in #models, which gets rebuilt from scratch on every refresh() call. The model selector triggers refresh("offline") on open, wiping all extension models from the list.
Store extension models as #runtimeModelOverlays that survive reloadStaticModels() and get merged during both #loadModels() and refreshRuntimeDiscoveries(). Persist extension API keys in a separate map restored after each clear cycle.
Remove unused buildCustomModel wrapper (sole call site refactored to use buildCustomModelOverlay + finalizeCustomModel directly).
- Added fallback logic to resolve stale chunk selectors by matching same-parent CRC when exact selector not found.
- Fixed stale child selector reuse to match chunks by checksum for sibling chunks with identical names.
- Added condition to check for Section chunk kind with Replace operation in region nullification logic.
- Expanded documentation with critical warnings about chunk name truncation and guidance for markdown editing with whitespace preservation.
- Added designer model role for UI/UX design tasks with Gemini 3.1 Pro as default model.
- Implemented model role fallback list support enabling automatic fallback to next available model when primary is unavailable.
- Refactored model role resolution to support multiple fallback patterns per role with thinking level mapping.
- Updated designer agent to use pi/designer role alias instead of explicit model list, removing spawns configuration.
- Added test coverage for model resolver fallback patterns and designer role override preferences.
- Added test coverage for geminiImageTool X-Title header routing through OpenRouter.
- Extracted tab width resolution from pi-natives to pi-utils with EditorConfig caching support.
- Made tabWidth parameter required in native text functions (visibleWidth, truncateToWidth, wrapTextWithAnsi, sliceWithWidth, extractSegments).
- Removed process-wide tab width state management from pi-natives text module.
- Consolidated wrapper node promotion logic in chunk classification with abbreviated string representations.
- Removed path-clean and pathdiff dependencies from pi-natives.
- Updated all consuming packages to import tab width functions from pi-utils and pass explicit tabWidth parameters.
