kon-coding-agent 0.3.4__tar.gz → 0.3.6__tar.gz

kon_coding_agent-0.3.6/.github/workflows/test.yml

@@ -0,0 +1,33 @@
+name: Test
+
+on: [push, pull_request]
+
+permissions:
+  contents: read
+
+jobs:
+  test:
+    runs-on: ubuntu-latest
+    strategy:
+      matrix:
+        python-version: ["3.12","3.13", "3.14"]
+    steps:
+    - uses: actions/checkout@v3
+    - name: Set up Python ${{ matrix.python-version }}
+      uses: actions/setup-python@v4
+      with:
+        python-version: ${{ matrix.python-version }}
+        cache: pip
+        cache-dependency-path: pyproject.toml
+    - name: Cache models
+      uses: actions/cache@v3
+      with:
+        path: ~/.cache
+        key: ${{ runner.os }}-torch-
+    - name: Install dependencies
+      run: |
+        pip install . --group dev
+    - name: Run tests
+      run: |
+        python -m pytest -s
+

{kon_coding_agent-0.3.4 → kon_coding_agent-0.3.6}/.kon/skills/kon-release-publish/SKILL.md

@@ -33,38 +33,44 @@ Use this skill when the user asks to cut a new Kon version, tag it, publish to P
    - `git status --short --branch` must be clean (or confirm with user)
    - `git tag --list` and `git log --oneline <prev_tag>..HEAD` to summarize changes
 
-2. **Version bump**
+2. **Update CHANGELOG.md**
+   - Replace the `## [Unreleased]` section's `- No changes yet.` with a new versioned heading: `## <version> - YYYY-MM-DD`
+   - Use `git log --oneline <prev_tag>..HEAD` to categorize changes into `### Added`, `### Changed`, `### Fixed` sections
+   - Credit external contributors with `- @username`
+   - Commit message: `docs: update changelog for <version>`
+
+3. **Version bump**
    - Update version in all 3 files above
 
-3. **Quality gates**
+4. **Quality gates**
    - `uv run ruff format .`
    - `uv run ruff check .`
    - `uv run pyright .`
    - `uv run pytest`
 
-4. **Commit**
+5. **Commit**
    - Commit message: `build: bump version to <version>`
 
-5. **Tag**
+6. **Tag**
    - Annotated tag: `git tag -a v<version> -m "v<version> ..."`
-   - Include concise “changes since previous tag” bullets
+   - Include concise "changes since previous tag" bullets
 
-6. **Push**
+7. **Push**
    - `git push origin main`
    - `git push origin v<version>`
 
-7. **Build + verify artifacts**
+8. **Build + verify artifacts**
    - `rm -rf dist && uv build`
    - `uv run python -m twine check dist/*`
 
-8. **Publish to PyPI**
+9. **Publish to PyPI**
    - Prefer token file if present (example `~/.pypi-token`):
    - `TWINE_USERNAME=__token__ TWINE_PASSWORD="$(< ~/.pypi-token)" uv run python -m twine upload dist/*`
    - Verify:
      - `https://pypi.org/project/kon-coding-agent/<version>/`
      - `https://pypi.org/pypi/kon-coding-agent/json` reports latest version
 
-9. **Create GitHub release**
+10. **Create GitHub release**
    - If token exists at `~/.github-token`, call Releases API:
    - `POST /repos/<owner>/<repo>/releases` with:
      - `tag_name: v<version>`
@@ -83,6 +89,7 @@ Use this skill when the user asks to cut a new Kon version, tag it, publish to P
 
 ## Output checklist to report
 
+- Changelog updated for `<version>`
 - Version bumped in all files
 - Checks passed
 - Commit hash

{kon_coding_agent-0.3.4 → kon_coding_agent-0.3.6}/AGENTS.md

@@ -4,6 +4,7 @@
 
 - Don't add trivial docstrings. Only add docstrings when explaining complex functionality.
 - This project uses `uv`. Run `uv run ruff format .` after editing or creating any files.
+- If generating and running a Python script, use `uv run python` instead of `python`.
 
 ## Testing
 

{kon_coding_agent-0.3.4 → kon_coding_agent-0.3.6}/CHANGELOG.md

@@ -6,6 +6,92 @@ All notable changes to this project will be documented in this file.
 
 - No changes yet.
 
+## 0.3.6 - 2026-04-23
+
+### Added
+
+- Configurable terminal bell notifications on response completion.
+- Render LaTeX math as Unicode in markdown output - @toojays.
+- Style approval popup with button-like controls and panel background.
+- Permission popup title card and improved UI formatting.
+
+### Changed
+
+- Improved UI for approval popup.
+- Reduce LaTeX preprocessing overhead.
+- Improved Codex SSE error extraction and retry on transient errors.
+
+### Fixed
+
+- Rewrite web_fetch to bypass bot-detection, preserve indentation, and harden against SSRF - @Meltedd (#28, #29).
+- Remove unused turn streaming state.
+- Remove unused tool imports.
+- Remove empty blocks.
+
+## 0.3.5 - 2026-04-18
+
+### Added
+
+- Standalone session HTML export with self-contained styling.
+- Configurable request timeout for API calls - @jspruit.
+- GitHub CI tests for Python 3.12–3.13 - @sukhbinder.
+- Width-aware popup lists and queue display.
+
+### Changed
+
+- Highlight color applied to the second column in floating lists.
+- Batch scroll during streaming, cache query_one lookups, pause spinner timer when idle.
+- Diff line length capped at 200 characters.
+
+### Fixed
+
+- Persist thinking level in session header and change all defaults to high.
+- Normalize OpenAI provider imports.
+- Widen resume popup labels.
+- Remove unused UI app import.
+
+### Performance
+
+- Added `gc.freeze()` and `PAUSE_GC_ON_SCROLL` to reduce GC stutters.
+
+## 0.3.4 - 2026-04-10
+
+### Fixed
+
+- Fixed Windows UTF-8 encoding errors in file operations - @sukhbinder.
+- Fixed local Gemma model thinking block compatibility.
+- Removed duplicate force-include for builtin skills in build config.
+
+### Changed
+
+- Updated local model documentation.
+
+## 0.3.3 - 2026-04-08
+
+### Added
+
+- Added queued agent steering between turns - @0xku.
+- Added bundled `/init` slash command for project scaffolding - @0xku.
+- Added help info for queue and steer queue commands.
+- Added GLM-5.1 support for zai provider.
+
+### Changed
+
+- Improved read tool directory listings.
+- Updated README with steer queue documentation.
+- Added `$` icon for bash, `%` for web tools, and `←` for edit tool.
+- Used muted color for shortcut key hints in exit/delete prompts.
+
+### Fixed
+
+- Let ESC interrupt retry backoff immediately - @0xku.
+- Fixed OpenAI login stdin leak by removing orphaned thread - @Meltedd.
+- Fixed OpenAI and Anthropic local compat with auth flags.
+- Fixed interrupt handling before handoff thread switch.
+- Fixed subprocess communication drain on cancellation/timeout.
+- Added zipfile path traversal validation.
+- Removed token throughput metrics.
+
 ## 0.3.2 - 2026-03-22
 
 ### Added

{kon_coding_agent-0.3.4 → kon_coding_agent-0.3.6}/PKG-INFO

@@ -1,19 +1,21 @@
 Metadata-Version: 2.4
 Name: kon-coding-agent
-Version: 0.3.4
+Version: 0.3.6
 Summary: Minimal coding agent
 License-File: LICENSE
 Requires-Python: >=3.12
 Requires-Dist: aiofiles>=25.1.0
 Requires-Dist: aiohttp>=3.13.3
 Requires-Dist: anthropic>=0.79.0
+Requires-Dist: curl-cffi>=0.15.0
 Requires-Dist: ddgs>=9.0.0
+Requires-Dist: html-to-markdown>=3.1.0
 Requires-Dist: openai>=2.21.0
 Requires-Dist: pillow>=12.1.1
 Requires-Dist: pydantic>=2.12.5
+Requires-Dist: readability-lxml>=0.8.4
 Requires-Dist: rich>=14.3.2
 Requires-Dist: textual>=8.0.0
-Requires-Dist: trafilatura>=2.0.0
 Description-Content-Type: text/markdown
 
 <h1 align="center">Kon</h1>
@@ -475,12 +477,6 @@ If `fd` or `rg` are missing, Kon can download them automatically.
 
 ---
 
-## Documentation
-
-- [Local models](docs/local-models.md) - Running Kon with local OpenAI-compatible models, plus tested setups and example commands.
-
----
-
 ## Acknowledgements
 
 - Kon takes significant inspiration from [pi coding-agent](https://github.com/badlogic/pi-mono/tree/main/packages/coding-agent), especially around philosophy and UI direction.

{kon_coding_agent-0.3.4 → kon_coding_agent-0.3.6}/README.md

@@ -457,12 +457,6 @@ If `fd` or `rg` are missing, Kon can download them automatically.
 
 ---
 
-## Documentation
-
-- [Local models](docs/local-models.md) - Running Kon with local OpenAI-compatible models, plus tested setups and example commands.
-
----
-
 ## Acknowledgements
 
 - Kon takes significant inspiration from [pi coding-agent](https://github.com/badlogic/pi-mono/tree/main/packages/coding-agent), especially around philosophy and UI direction.

kon_coding_agent-0.3.6/docs/code-health-scan.md

@@ -0,0 +1,144 @@
+# Code Health Scan: Additional Notable Issues
+
+This note captures only the additional issues found in a fresh repo scan that are **not** the same as the already-known architectural concerns in `docs/architecture-review.md`.
+
+The baseline conclusion from the earlier review still stands:
+
+- the core runtime is strong
+- the TUI still owns too much runtime/session orchestration
+- the large controller/runtime-manager refactor is deferred for now
+
+This document is intentionally narrower. It records only the standout issue and the smaller cleanup opportunities found in the current repo state after the recent cleanup pass.
+
+## Overall verdict
+
+Aside from the already-documented UI/runtime orchestration boundary issue, the repo looks to be in decent shape overall.
+
+Notable signals:
+
+- test suite passed cleanly
+- pyright passed cleanly
+- only a few lint issues remained
+
+So this is **not** a case where the repo appears to have multiple additional architectural problems lurking elsewhere. The main new thing worth tracking is a state-consistency problem around provider/model transitions, plus a few smaller cleanup items.
+
+## Major issue
+
+### Provider/model transition paths are not transactional
+
+The main additional issue is that some provider/model/session transition flows update app state before it is certain that the replacement provider/runtime state can be constructed successfully.
+
+Relevant files:
+
+- `src/kon/ui/commands.py`
+- `src/kon/ui/session_ui.py`
+
+### Where it shows up
+
+#### `src/kon/ui/commands.py::_select_model`
+
+This path updates:
+
+- `self._model`
+- `self._model_provider`
+
+before provider recreation is guaranteed to succeed.
+
+If `_create_provider(...)` fails, the UI reports the error, but some selected-model state may already have been changed while the active provider/agent runtime is still using the old provider instance.
+
+#### `src/kon/ui/session_ui.py::_load_session`
+
+This path loads session metadata and updates model/provider-related state while also trying to reconcile the active provider.
+
+If provider recreation fails during session load, the code reports the error, but can still leave partially updated state behind:
+
+- loaded session state may now be active
+- selected model/provider values may reflect the resumed session
+- the provider instance may still be the old one or only partially reconfigured
+
+### Why this matters
+
+This is not a large-architecture problem, but it is a real correctness risk.
+
+It can produce subtle mismatches between:
+
+- selected model/provider UI state
+- active provider config
+- agent runtime state
+- resumed session metadata
+
+Those failures are harder to reason about than a simple hard error because the app can continue running in a partially updated state.
+
+### Suggested direction
+
+Without doing the large architecture refactor, the safer short-term fix would be to make these transitions more atomic:
+
+1. compute the target model/provider/session state first
+2. build or validate the replacement provider first
+3. only commit the new state to `self._model`, `self._model_provider`, `self._provider`, `self._session`, and `self._agent` after success
+4. otherwise leave the old runtime state untouched
+
+Even a small helper that stages the new state before assignment would reduce the sync-risk significantly.
+
+## Small cleanup opportunities
+
+### 1. `CompactionEntry.first_kept_entry_id` is currently misleading
+
+Relevant file:
+
+- `src/kon/session.py`
+
+`CompactionEntry` stores `first_kept_entry_id`, and comments imply that the compacted view depends on it.
+
+However, `Session.messages` currently reconstructs the compacted view by:
+
+- finding the last compaction entry
+- inserting the synthetic summary pair
+- including message entries after the compaction entry itself
+
+It does **not** currently use `first_kept_entry_id` to decide what to retain.
+
+This is not a functional bug in current usage because existing callers/tests append compaction entries in a way that still makes the behavior correct. But the data model and implementation intent are slightly out of sync, which makes the field more confusing than helpful right now.
+
+Small fix options:
+
+- actually use `first_kept_entry_id` in `Session.messages`, or
+- simplify the field/comment contract if the compaction-entry position is the real source of truth
+
+### 2. System-prompt fallback logic is still duplicated across UI mixins
+
+Relevant files:
+
+- `src/kon/ui/commands.py`
+- `src/kon/ui/session_ui.py`
+
+Both files define the same `_resolve_system_prompt(...)` helper:
+
+- use persisted `session.system_prompt` when available
+- otherwise rebuild with `build_system_prompt(...)`
+
+This is a small issue, not a major design problem, but it means the recent centralization is still only partial. If the fallback behavior changes again, both copies will need to stay in sync.
+
+Small fix option:
+
+- move the helper to one shared non-duplicated location used by both mixins/app paths
+
+### 3. Lint cleanup remains in a few tool modules
+
+Relevant files:
+
+- `src/kon/tools/find.py`
+- `src/kon/tools/grep.py`
+- `src/kon/tools/web_fetch.py`
+
+Current lint output shows unused imports of `truncate_text` in these files.
+
+This is only a hygiene issue, but it is a useful signal that a full lint pass likely did not happen after some recent edits.
+
+## Final assessment
+
+The repo looks fine overall.
+
+Aside from the already-known concerns in `docs/architecture-review.md`, the only additional issue that really stands out is the non-transactional provider/model/session transition behavior.
+
+The other findings are small cleanup opportunities rather than signs of deeper architectural trouble.

{kon_coding_agent-0.3.4 → kon_coding_agent-0.3.6}/pyproject.toml

@@ -14,7 +14,7 @@ default = true
 
 [project]
 name = "kon-coding-agent"
-version = "0.3.4"
+version = "0.3.6"
 description = "Minimal coding agent"
 readme = "README.md"
 requires-python = ">=3.12"
@@ -22,13 +22,15 @@ dependencies = [
     "aiofiles>=25.1.0",
     "aiohttp>=3.13.3",
     "anthropic>=0.79.0",
+    "curl-cffi>=0.15.0",
+    "ddgs>=9.0.0",
+    "html-to-markdown>=3.1.0",
     "openai>=2.21.0",
     "pillow>=12.1.1",
     "pydantic>=2.12.5",
+    "readability-lxml>=0.8.4",
     "rich>=14.3.2",
     "textual>=8.0.0",
-    "ddgs>=9.0.0",
-    "trafilatura>=2.0.0",
 ]
 
 [dependency-groups]
@@ -60,6 +62,10 @@ select = [
     "RUF", # ruff-specific rules
 ]
 
+[tool.ruff.lint.per-file-ignores]
+"src/kon/ui/latex.py" = ["RUF001"]
+"tests/ui/test_latex.py" = ["RUF001"]
+
 [tool.ruff.lint.isort]
 split-on-trailing-comma = false
 

{kon_coding_agent-0.3.4 → kon_coding_agent-0.3.6}/src/kon/config.py

@@ -76,6 +76,7 @@ class LLMConfig(BaseModel):
     default_thinking_level: str
     system_prompt: SystemPromptConfig
     tool_call_idle_timeout_seconds: float = 180
+    request_timeout_seconds: float = 600
     auth: AuthConfig = AuthConfig()
 
 
@@ -97,6 +98,10 @@ class ToolsConfig(BaseModel):
     extra: list[str] = []
 
 
+class NotificationsConfig(BaseModel):
+    enabled: bool = False
+
+
 class ConfigSchema(BaseModel):
     meta: MetaConfig
     llm: LLMConfig
@@ -105,6 +110,7 @@ class ConfigSchema(BaseModel):
     agent: AgentConfig
     tools: ToolsConfig = ToolsConfig()
     permissions: PermissionsConfig
+    notifications: NotificationsConfig = NotificationsConfig()
 
 
 # =================================================================================================
@@ -193,6 +199,10 @@ class Config:
     def tools(self) -> ToolsConfig:
         return self._parsed.tools
 
+    @property
+    def notifications(self) -> NotificationsConfig:
+        return self._parsed.notifications
+
     @property
     def binaries(self) -> _BinariesConfig:
         return _BinariesConfig(AVAILABLE_BINARIES)

{kon_coding_agent-0.3.4 → kon_coding_agent-0.3.6}/src/kon/defaults/config.toml

@@ -11,6 +11,9 @@ default_thinking_level = "high"
 # Abort a tool call if it stays idle for this long during a turn.
 # Helps prevent stalled tool executions from hanging the agent loop.
 tool_call_idle_timeout_seconds = 180
+# HTTP request timeout for LLM API calls (in seconds).
+# Local models (e.g. llama.cpp) may need a higher value for long compaction requests.
+request_timeout_seconds = 600
 
 [llm.auth]
 # Auth policy for OpenAI-compatible and Anthropic-compatible endpoints.
@@ -70,3 +73,8 @@ collapse_thinking = true
 # "prompt" asks before edits/writes and other mutating actions.
 # "auto" allows tool calls without approval prompts.
 mode = "prompt" # "prompt" or "auto"
+
+[notifications]
+# Ring the terminal bell when kon finishes a task or waits for tool approval.
+# Disabled by default.
+enabled = false

{kon_coding_agent-0.3.4 → kon_coding_agent-0.3.6}/src/kon/llm/base.py

@@ -77,7 +77,7 @@ class ProviderConfig:
     model: str = ""
     max_tokens: int = 8192
     temperature: float | None = None
-    thinking_level: str = "medium"
+    thinking_level: str = "high"
     provider: str | None = None
     session_id: str | None = None
     openai_compat_auth_mode: AuthMode = "auto"

{kon_coding_agent-0.3.4 → kon_coding_agent-0.3.6}/src/kon/llm/providers/openai_codex_responses.py

@@ -279,11 +279,26 @@ class OpenAICodexResponsesProvider(BaseProvider):
                             yield StreamDone(stop_reason=stop_reason)
                             return
 
-                    elif event_type in {"response.failed", "error"}:
+                    elif event_type == "error":
+                        code = event.get("code")
                         message = event.get("message")
-                        if not isinstance(message, str):
-                            message = "Codex response failed"
-                        yield StreamError(error=message)
+                        if isinstance(message, str) and message:
+                            yield StreamError(error=f"Codex error: {message}")
+                        elif isinstance(code, str) and code:
+                            yield StreamError(error=f"Codex error: {code}")
+                        else:
+                            yield StreamError(error=f"Codex error: {json.dumps(event)}")
+                        return
+
+                    elif event_type == "response.failed":
+                        response_obj = event.get("response")
+                        msg = None
+                        if isinstance(response_obj, dict):
+                            err = response_obj.get("error")
+                            if isinstance(err, dict):
+                                msg = err.get("message")
+                        err_msg = msg if isinstance(msg, str) and msg else "Codex response failed"
+                        yield StreamError(error=err_msg)
                         return
             finally:
                 await session.close()
@@ -321,7 +336,8 @@ class OpenAICodexResponsesProvider(BaseProvider):
         return StopReason.STOP
 
     def should_retry_for_error(self, error: Exception) -> bool:
-        return False
+        msg = str(error).lower()
+        return any(kw in msg for kw in ("429", "rate_limit", "server_error", "502", "503", "504"))
 
 
 def is_openai_logged_in() -> bool:

{kon_coding_agent-0.3.4 → kon_coding_agent-0.3.6}/src/kon/llm/providers/openai_completions.py

@@ -10,6 +10,8 @@ from openai.types.chat import (
     ChatCompletionToolParam,
 )
 
+from kon import config as kon_config
+
 from ...core.types import (
     AssistantMessage,
     ImageContent,
@@ -99,7 +101,11 @@ class OpenAICompletionsProvider(BaseProvider):
                 "Set OPENAI_API_KEY or ZAI_API_KEY environment variable, "
                 'or configure llm.auth.openai_compat = "auto"/"none" for local endpoints.'
             )
-        self._client = AsyncOpenAI(api_key=api_key, base_url=config.base_url)
+        self._client = AsyncOpenAI(
+            api_key=api_key,
+            base_url=config.base_url,
+            timeout=kon_config.llm.request_timeout_seconds,
+        )
         self._compat = _detect_compat(
             config.provider or "", config.base_url or "", config.model or ""
         )

{kon_coding_agent-0.3.4 → kon_coding_agent-0.3.6}/src/kon/llm/providers/openai_responses.py

@@ -4,6 +4,8 @@ from typing import Any
 
 from openai import APIStatusError, AsyncOpenAI, RateLimitError
 
+from kon import config as kon_config
+
 from ...core.types import (
     AssistantMessage,
     ImageContent,
@@ -51,6 +53,7 @@ class OpenAIResponsesProvider(BaseProvider):
                 api_key=self.config.api_key,
                 base_url=self.config.base_url,
                 default_headers=self._headers,
+                timeout=kon_config.llm.request_timeout_seconds,
             )
         return self._client
 

kon_coding_agent-0.3.6/src/kon/notify.py

@@ -0,0 +1,30 @@
+import os
+import time
+
+_BELL_DEBOUNCE_S = 0.5
+_last_bell_time: float = 0.0
+
+
+def _raw_write(data: bytes) -> None:
+    try:
+        fd = os.open("/dev/tty", os.O_WRONLY | os.O_NOCTTY)
+        try:
+            os.write(fd, data)
+        finally:
+            os.close(fd)
+    except OSError:
+        os.write(2, data)
+
+
+def _bell() -> None:
+    global _last_bell_time
+    now = time.monotonic()
+    if now - _last_bell_time < _BELL_DEBOUNCE_S:
+        return
+    _last_bell_time = now
+    _raw_write(b"\a")
+
+
+def notify(title: str, message: str) -> None:
+    del title, message
+    _bell()