HTML good first issues

AI Summary: This issue is a placeholder to direct users to the official Discord server for general discussions, support, and help related to EaseMotion CSS. It clarifies that bugs and feature proposals should use separate templates. The issue itself doesn't present a technical problem but rather a communication channel.

AI Summary: This issue identifies a minor factual error in the `corticospinal_tract_axonopathy` module's notes on GitHub. The notes incorrectly associate the gene GJC2 with SPG42; GJC2 is actually SPG44. The fix involves correcting the locus number from SPG42 to SPG44 in the documentation to ensure accurate disease routing for curators.

AI Summary: Four Nx skills within the full-stack-app are unattributed stock content that conflict with the core's closed-set generator model. The issue requires a decision on how to handle these skills, either by attributing them, replacing them with core-specific implementations, or removing them.

AI Summary: Documentation in `CLAUDE.core.md` and `README.md` files across several core packages incorrectly describes the build loop as using `hedgehog next`. In reality, the loop skills use `hedgehog claim`. This issue requires auditing all relevant documentation files to ensure they accurately reflect the actual command sequence used by the loop skills, likely by including both `next` and `claim` commands.

AI Summary: The bootstrap process for authored core projects incorrectly attempts to run verification commands that contain an unsubstituted `{module}` placeholder. This occurs because the substitution for `{module}` happens during the `hedgehog plan` compilation, which is executed after the bootstrap step. The issue proposes fixing this by either substituting a placeholder module name during bootstrap or by explicitly stating that module-axis layer verification commands cannot run as-is at this stage.

AI Summary: This issue involves auditing the skills and agents within the `hedgehog-core-adopted` repository. The goal is to identify and document defects in the content and logic of these components, which are responsible for adopting existing codebases. Findings will be addressed through small pull requests to the `hedgehog-core-adopted` repository.

AI Summary: This issue is part of a larger effort to audit the 'authored core' of a project, specifically focusing on its skills and agents. The goal is to identify defects in the content of the `hedgehog-core-authored` repository by performing a manual review, documenting findings, and submitting small fixes. The audit will look for logical inconsistencies, verbosity, and outdated information within the specified skills and agents.

AI Summary: This issue involves auditing the skills and agents within the `deepseek-harness` core repository. The goal is to identify and document defects in the content and logic of these components, specifically focusing on their interaction with the DeepSeek Harness runtime. The findings will be used to create small pull requests to fix issues in the `deepseek-harness` repository.

AI Summary: This issue is a code audit of the landing-page core repository, focusing on its skills and agents. The goal is to identify and document defects related to logic, verbosity, and unnecessary ceremony, with the intention of creating small pull requests to fix them. Findings related to shared engine code or the core.yaml contract are out of scope.

AI Summary: This issue is a code audit of the 'pwa-app' core, focusing on its skills and agents. The goal is to identify and document defects related to logic, verbosity, unnecessary ceremony, and over-engineering within specific skills and agents. The findings will be used to create small pull requests against the 'hedgehog-core-pwa-app' repository.

AI Summary: This issue is part of a larger effort to audit and improve the skills and agents within the `full-stack-app` core of the Hedgehog project. The goal is to identify and fix issues related to logic, verbosity, unnecessary ceremony, and over-engineering in the specified skills and agents. The findings will be documented and addressed through small pull requests to the `hedgehog-core-full-stack-app` repository.

AI Summary: This issue is part of a larger effort to audit and refactor the core engine's skills and agents. The goal is to identify and fix issues related to logic, verbosity, historical leakage, and unnecessary ceremony within the agent and skill content. The audit will involve a manual review of specified files, followed by small pull requests to address identified defects.

AI Summary: This issue proposes adding a new crafting recipe for a "Skull Mask" to the game. The user suggests "Rope + Skull" as potential ingredients, and the item's internal name is identified as WZ_SkullMask_Plain. This is labeled as an enhancement and a good first issue.

AI Summary: This issue aims to audit the `.hedgehog/core.yaml` contract from both the code parsing side and the agent/skill interpretation side. The goal is to identify semantic discrepancies where agents or skills interpret fields differently than the code that parses and uses them, or where agents expect fields that core code doesn't emit.

AI Summary: This issue proposes a comprehensive audit of the skill and agent content within the Hedgehog project. The goal is to identify and rectify inconsistencies, redundancies, and outdated information that have accumulated over time, leading to inefficiencies and potential errors. The audit will cover both the engine and various core components, resulting in a defect list and subsequent pull requests for fixes.

AI Summary: This issue addresses unmet peer dependencies within the `full-stack-app` workspace, specifically concerning ESLint, Rspack, and Noble Hashes. The goal is to identify actionable dependencies, update them to compatible versions, or implement overrides if necessary, ensuring the workspace's linting layer gate remains functional.

AI Summary: This issue proposes adding an automated check to ensure the explainer count is consistent across four documentation files: README.md, CONTRIBUTORS.md, METRICS.md, and ROADMAP.md. Currently, manual updates can lead to discrepancies, and the existing checks only verify generated site artifacts. The suggested fix is a script that counts explainer files and verifies this count against specific mentions in the mentioned documentation files, integrated into the CI workflow.

AI Summary: The web profiler's Compare tool is missing the 'Advanced thresholds' panel, which is present in the single-dataset Profile view and the CLI. This prevents users from customizing thresholds like `min_share` and `min_group_size` when comparing datasets, unlike other profiler surfaces. The suggested fix involves adding the same threshold input panel to the Compare view and passing the selected options through the comparison function.

AI Summary: The Tenant Screening project is missing from two mapping dictionaries, `PROJECT_ANCHORS` and `projectAnchors`, which are used to generate internal links within explainer pages. This prevents links to the Tenant Screening audit from being correctly converted to anchor links, causing them to remain as plain folder links. The fix involves adding the missing entry to both maps and removing a dead, misspelled legacy key from the JavaScript map.

AI Summary: The MCP server's `compare_datasets` tool currently lacks support for proxy hints, which are available in the CLI version and the `profile_dataset` tool. This enhancement would involve adding a `proxy_hints` parameter to `compare_datasets` and implementing the necessary logic to compute and handle proxy hints for both datasets being compared.

AI Summary: The MCP server's proxy_hints tool currently lacks a way to pass `held_out` data, which is a feature already implemented in the underlying `proxy_hints` function and exposed via the CLI. This enhancement would involve adding a parameter to the MCP tool to mirror the CLI's `--proxy-hints-with` functionality, allowing agents to test protected attributes that have been removed from the dataset.

AI Summary: The `faircode profile` command incorrectly handles the `--proxy-hints-with` flag. If `--proxy-hints` is not provided, `--proxy-hints-with` is parsed but its value is ignored, leading to silent failures when the specified file doesn't exist. The suggested fix is to either explicitly error when `--proxy-hints` is missing or to automatically imply `--proxy-hints` when `--proxy-hints-with` is used.

AI Summary: This issue proposes adding a new explainer document to the repository that addresses the problem of multiple comparisons and the False Discovery Rate (FDR) in fairness audits. It highlights that running numerous statistical tests without correction can lead to an inflated number of false positives, and suggests explaining various correction methods like Bonferroni, Holm, and Benjamini-Hochberg FDR with concrete examples and runnable code.

AI Summary: This issue proposes adding a new explainer document to the repository that details the fairness-accuracy trade-off in machine learning models. It should explain why improving fairness often comes at the cost of accuracy, using concrete examples and data from existing CSV files to illustrate the concept and demonstrate mitigation strategies.

AI Summary: This issue requests the addition of a new explainer document for the 'Treatment Equality' fairness metric. The author notes that existing explainers in the repository do not cover this specific metric, which focuses on the ratio of false negatives to false positives being equal across groups. The suggested structure for the new explainer is provided, along with a note that the metric is not currently computed by the repository's benchmark script.

AI Summary: The `check_em_dash.py` script currently fails to scan `.yml` and `.yaml` files for em dashes. This is a bug because the repository's style guide applies to all files, and two existing `.yml` files contain em dashes for demonstration purposes without being flagged. The suggested fix is to include these file extensions in the scan and add the specific demonstration files to the allowlist.

AI Summary: The `llms.txt` file, which serves as a short index for AI answer engines, is missing the 6 most recently added explainers. This makes these explainers invisible to crawlers that only access the index file. The issue suggests manually adding the missing entries and potentially updating the CI checks to prevent future drift.

AI Summary: The roadmap grid on the `index.html` page incorrectly displays three published explainers as 'coming soon'. These explainers have corresponding files and entries in the data JSON, but their status class was never updated from 'soon' to 'done'. The suggested fix is to change the status class for these three items to 'done' to accurately reflect their published status.

AI Summary: This issue highlights a classification problem in the MONDO ontology where Fanconi anemia complementation group A (FA-A) is incorrectly classified as a skeletal dysplasia, while its parent and sibling terms are not. The author argues this is biologically inaccurate and points to a poorly defined intermediate term in the ontology's hierarchy. Additionally, the author disputes Fanconi anemia's classification as an inborn error of metabolism.

AI Summary: The documentation in CLAUDE.md for pre-PR gates is incomplete, omitting four essential checks that `just qc` performs. This omission led to a bug where a snippet was truncated mid-word, which was only caught by one of the missing gates. The suggested fix is to update the documentation to include all relevant gates or to dynamically list the output of `just qc`.