AI/ML good first issues

AI Summary: This issue focuses on completing the Project Continuity feature by implementing automatic crash detection and a guided restart-recovery workflow. The goal is to allow users to resume their work after an unexpected interruption without manually copying sensitive provider sessions or fabricating progress, using persisted project state as the source of truth.

AI Summary: This issue proposes adding a native Pi extension that integrates with Perenna via a shared CLI path. The extension will act as an adapter, registering memory-related tools with Pi and forwarding requests to the Perenna CLI without reimplementing Perenna's logic. This integration aims to simplify Pi's interaction with Perenna by bypassing the MCP layer.

AI Summary: This issue proposes adding a new CLI adapter to Perenna that reuses the existing MCP memory command path for memory operations. This will allow clients to interact with memories (list, search, get, create, patch, delete) via the CLI without duplicating logic and risking inconsistencies with the MCP tools.

AI Summary: This issue highlights a critical problem where the pack's installation can fail silently at multiple stages, leaving users unaware of broken functionality. The proposed solution is a read-only diagnostic script that checks and reports the status of agent visibility, skill paths, routing blocks, and hook wiring in plain language, with clear next steps for any issues.

AI Summary: This issue identifies that three skills within the project lack a "worked example" in their respective `references/` directories, as required by the `CONTRIBUTING` guidelines. The task is to create a synthetic or self-owned worked example for each of these three skills, ensuring they adhere to the format of existing examples and the rules outlined in their `SKILL.md` files. The examples must not contain confidential or third-party data.

AI Summary: This issue proposes refactoring the scoring logic in `evals/score.py` to avoid hardcoding model tiers. The goal is to allow new model tiers to be added by simply updating a registry, rather than modifying the code, and to ensure consistent ordering and display names for models.

AI Summary: The user is experiencing a significant slowdown (50-60 seconds) in app startup on real Android devices with versions v1.5.x, compared to v1.4.0 (10 seconds) and simulators (4 seconds). They have ruled out several common causes like external storage, update checks, network issues, and payload differences. The core problem appears to be a slow engine (node) readiness after decompression on arm64 devices, which is puzzling as v1.4.0 uses the same kernel and node but starts much faster.

AI Summary: This issue requests the implementation of UI localization (i18n) to support languages other than English, specifically mentioning Russian. The current UI has all strings hardcoded, making it difficult for non-English speakers who rely on unreliable browser auto-translation. The proposed solution involves creating an i18n layer for UI strings and adding a language switcher.

AI Summary: This issue proposes adding configurable frequency controls for the `autoDistill` feature, which currently triggers a distillation reminder after every tool-use turn. The goal is to allow users to reduce the frequency of these reminders by setting parameters for turn interval, time cooldown, or a minimum number of tool calls per turn, while maintaining the core functionality of automatic distillation.

AI Summary: This feature request proposes migrating the "Memory Panel" from the settings page to the main interface's sidebar. The goal is to improve accessibility for this frequently used feature by making it a permanent fixture and displaying it in a right-side drawer to maintain context with the main conversation. The proposed design prioritizes search and filtering over scrolling for better usability with a large number of memories.

AI Summary: This feature request proposes a system to import conversation history from various AI coding tools (like Claude Code, Codex) into a unified memory database. It suggests an adapter pattern to parse different file formats, an LLM to extract structured memories (preferences, decisions, knowledge), and crucial deduplication mechanisms to prevent redundancy. The implementation would include manual preview and selection of imported memories, along with CLI and GUI triggers.

AI Summary: This issue asks for verification of the zero-cost Docker Compose setup process on macOS or Linux. The goal is to ensure a smooth first-time experience for users by documenting any friction encountered during a fresh checkout and setup.

AI Summary: The maintainers of Personal AI OS are seeking genuine external user feedback on real-world workflow problems the OS solves. They are asking users to share their experiences, including what worked, what didn't, and what improvements would be most impactful, to validate the project's utility beyond the maintainer's own use.

AI Summary: A user on macOS 13 (Ventura) with an Apple Silicon Mac is encountering a "program is damaged" error when trying to open an application. The issue is marked as a duplicate, help wanted, and won't fix, suggesting it's a known problem that may not be actively addressed or has been resolved elsewhere.

AI Summary: The model picker incorrectly displays deprecated OpenAI models that result in a 404 error when used. This issue arises because the system builds the picker directly from an API response that still lists these defunct models. The proposed solutions involve either a deny-list, probing models to cache their availability, or marking unavailable models in the UI.

AI Summary: This issue proposes to refactor the `test_response_quality.py` file within the `evals/judge/` directory. Currently, it uses a `criteria`-based approach for evaluation, which has shown inconsistencies. The goal is to migrate this to an `evaluation_steps` convention, similar to `test_retrieval_gate_accuracy.py`, to ensure greater reproducibility and a stable evaluation rubric. This involves rewriting the existing criteria into explicit steps and verifying the impact on evaluation scores.

AI Summary: This issue proposes adding a new "deno" stack preset to the project. This involves creating configuration files for Deno's code checking, linting, and testing within the pre-commit hook, and updating documentation to reflect the new preset.

AI Summary: This issue requests the addition of a new "php-laravel" stack preset to the claude-scaffold project. This involves creating a rules file and a pre-commit hook script, along with mirroring the rules for English documentation and updating the stack table in the README.

AI Summary: This issue requests the addition of a new "dotnet" stack preset to the project. This involves creating a rules file and a pre-commit hook script that will run `dotnet build`, `dotnet format --verify-no-changes`, and `dotnet test`. The contributor needs to follow the provided documentation for creating new presets and update the README.

AI Summary: The WhaleHarness audit of ccch1mneyyy/dsh-TUI@0.5.3 was rejected because the patch loads foreign packages from '@deepseek-ai'. The issue indicates that most rejections are easily fixable and provides a link to a guide for common fixes.

AI Summary: The dsh-TUI terminal frontend lacks a dedicated UI for configuring plugins. Currently, users can only adjust settings by manually editing YAML files or using a basic `/provider` wizard, which is insufficient for plugins like `dsh-advisor` that have their own settings namespaces.

AI Summary: This issue discusses whether the `lib/` directory, containing build artifacts, should be committed to Git. While currently included, it causes review and merge conflicts. The author proposes adding a `prepare` script to automatically build on install, allowing `lib/` to be ignored by Git, but notes this might affect installations via Git URLs.

AI Summary: This issue asks contributors to test the "zero-cost" setup path for Personal AI OS on Windows, meaning users should be able to run the application without needing an API key. The goal is to ensure the documented setup process is clear and functional for new Windows users and to identify any friction points that automated CI might miss.

AI Summary: This issue proposes integrating advanced features into a tool (tui) for professional users in scientific research. Key requirements include seamless SSH deployment to remote Linux servers and cloud GPUs, robust management of long-running multi-CPU/GPU tasks, and a plugin system for workflow and configuration cloning. The goal is to establish tui as a foundational element within a larger remote SSH ecosystem (dsh).

AI Summary: This issue requests early user feedback on the Personal AI OS `main` branch before the `v0.2.0` release. Testers are asked to perform a fresh install, configure a provider without exposing credentials, and complete a basic chat interaction, reporting any issues encountered during setup or usage.

AI Summary: This issue requests full internationalization (i18n) support for the OptiQra application, specifically addressing the audit rules which are currently not localized. Implementing this would allow users worldwide to fully utilize and understand the application in their native language.

AI Summary: This issue proposes adding a new performance audit to detect whether a scanned website is served over HTTP/2 or HTTP/3. This information will be presented as a performance signal, as these newer protocols offer significant load-time improvements compared to HTTP/1.1. The implementation will involve modifying the performance audit module and potentially the API route responsible for outbound fetches.

AI Summary: This issue proposes enhancing the report UI to visually represent Core Web Vitals (CWV) data, moving beyond plain numbers. The goal is to implement visual indicators like gauges or bar charts for LCP, CLS, and INP, mirroring the user-friendly presentation of tools like PageSpeed Insights. This will make CWV metrics more scannable and understandable for developers.

AI Summary: This issue is a call for three open-source maintainers to participate in a pilot program for a read-only Repo Steward AI workflow. Participants will run the AI on their repositories for 7-14 days, performing tasks like pull-request review and issue triage, and will provide feedback through a pilot report. The goal is to gather data on the AI's effectiveness and identify areas for improvement.