Conversation
Adds Garry Tan's gstack open-source engineering toolkit as a project skill collection under .claude/skills/gstack, enabling slash commands like /office-hours, /review, /qa, /ship, and more. Creates CLAUDE.md documenting the available skills. https://claude.ai/code/session_01RFRc5VXgSqsuBZGki1aaQG
When the player enters DOWNHILL, nearby bears (within 1 screen height above, 2 screen widths laterally) automatically join the race and steer toward the player to jostle them — bumpercars style. Bears continuously enroll as the player descends past their altitude. Race features: - Bumpercars AI: gravity + lateral steering toward player X, bump impulse on collision - Bear radar: screen-edge arrow indicators pointing at off-screen racing bears, fading in as they get closer - Force-spawn guarantee: if 3 seconds pass with no racing bear nearby, one spawns just above the visible screen so the player always has competition - Race result moment: win/lose/mixed floating text with money bonus ($100 per bear beaten), color-coded green/gold/brown - Scatter animation: on race end, bears dash off screen at speed before snapping back to their home layer — no instant teleport - Stamina fix: game now properly starts in UPHILL mode; reaching the peak auto-transitions to DOWNHILL; Tab still works anywhere as manual toggle Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
✅ Deploy Preview for sledhead ready!
To edit notification comments on pull requests, go to your Netlify project configuration. |
There was a problem hiding this comment.
Pull request overview
This PR vendors and expands the gstack skill set and its browse (Playwright-backed) headless browser daemon, adding new skills/docs plus substantial new functionality and test coverage around navigation security, cookie importing, snapshotting, and Windows compatibility.
Changes:
- Add new skills/templates (e.g.,
/careful,/canary,/benchmark,/browse) and generated docs. - Implement/extend
browseserver + command set (URL validation, path validation, cookie picker routes, snapshot flags, buffers, Windows Node fallback polyfill). - Add a large Bun test suite with fixtures to validate the new browse behaviors end-to-end.
Reviewed changes
Copilot reviewed 77 out of 235 changed files in this pull request and generated 8 comments.
Show a summary per file
| File | Description |
|---|---|
| .claude/skills/gstack/careful/SKILL.md.tmpl | Adds /careful skill template with destructive-command guardrails |
| .claude/skills/gstack/careful/SKILL.md | Generated /careful skill doc from template |
| .claude/skills/gstack/canary/SKILL.md.tmpl | Adds /canary post-deploy monitoring skill template |
| .claude/skills/gstack/browse/test/url-validation.test.ts | Adds tests for dangerous scheme + metadata endpoint blocking |
| .claude/skills/gstack/browse/test/test-server.ts | Adds tiny Bun test server for HTML fixtures |
| .claude/skills/gstack/browse/test/platform.test.ts | Adds tests for platform constants + isPathWithin |
| .claude/skills/gstack/browse/test/path-validation.test.ts | Adds tests for output/read path validation helpers |
| .claude/skills/gstack/browse/test/handoff.test.ts | Adds tests for handoff/resume and state persistence |
| .claude/skills/gstack/browse/test/gstack-config.test.ts | Adds tests for bin/gstack-config script behavior |
| .claude/skills/gstack/browse/test/fixtures/upload.html | Adds upload fixture page |
| .claude/skills/gstack/browse/test/fixtures/states.html | Adds element-states fixture page |
| .claude/skills/gstack/browse/test/fixtures/spa.html | Adds SPA fixture with console output |
| .claude/skills/gstack/browse/test/fixtures/snapshot.html | Adds snapshot fixture with mixed interactive elements |
| .claude/skills/gstack/browse/test/fixtures/responsive.html | Adds responsive layout fixture page |
| .claude/skills/gstack/browse/test/fixtures/qa-eval.html | Adds QA-eval fixture with intentionally planted bugs |
| .claude/skills/gstack/browse/test/fixtures/qa-eval-spa.html | Adds SPA QA-eval fixture with intentionally planted bugs |
| .claude/skills/gstack/browse/test/fixtures/qa-eval-checkout.html | Adds checkout QA-eval fixture with intentionally planted bugs |
| .claude/skills/gstack/browse/test/fixtures/forms.html | Adds forms fixture page |
| .claude/skills/gstack/browse/test/fixtures/empty.html | Adds minimal empty-page fixture |
| .claude/skills/gstack/browse/test/fixtures/dialog.html | Adds dialog fixture page |
| .claude/skills/gstack/browse/test/fixtures/cursor-interactive.html | Adds cursor-interactive fixture page |
| .claude/skills/gstack/browse/test/fixtures/basic.html | Adds basic content fixture page |
| .claude/skills/gstack/browse/test/find-browse.test.ts | Adds tests for locating the compiled browse binary |
| .claude/skills/gstack/browse/test/cookie-picker-routes.test.ts | Adds route-level tests for cookie picker JSON/CORS behavior |
| .claude/skills/gstack/browse/test/config.test.ts | Adds tests for config resolution, gitignore updates, health checks, etc. |
| .claude/skills/gstack/browse/test/bun-polyfill.test.ts | Adds tests for Node Bun polyfill behavior |
| .claude/skills/gstack/browse/src/write-commands.ts | Adds/extends write commands (click routing, upload, cookie import, cookie picker open) |
| .claude/skills/gstack/browse/src/url-validation.ts | Adds URL validation (scheme blocking + metadata endpoint protection + DNS rebinding guard) |
| .claude/skills/gstack/browse/src/snapshot.ts | Implements snapshot flags parsing, cursor-interactive scan, annotate/diff support |
| .claude/skills/gstack/browse/src/server.ts | Adds/extends browse server (auth, buffers flush, help text, health route, cookie picker routing) |
| .claude/skills/gstack/browse/src/read-commands.ts | Adds/extends read commands (path validation, is state checks, secret redaction) |
| .claude/skills/gstack/browse/src/platform.ts | Adds cross-platform TEMP_DIR + path-within helper |
| .claude/skills/gstack/browse/src/meta-commands.ts | Adds meta commands (screenshot/pdf/responsive, chain, diff, handoff/resume) + output path validation |
| .claude/skills/gstack/browse/src/find-browse.ts | Adds binary discovery tool for browse executable |
| .claude/skills/gstack/browse/src/cookie-picker-routes.ts | Adds cookie picker HTTP route handler (UI + JSON API) |
| .claude/skills/gstack/browse/src/config.ts | Adds shared config resolution + .gstack/ state dir + .gitignore integration |
| .claude/skills/gstack/browse/src/commands.ts | Adds central command registry + descriptions with load-time validation |
| .claude/skills/gstack/browse/src/bun-polyfill.cjs | Adds Node polyfill for Bun APIs (Windows fallback) |
| .claude/skills/gstack/browse/src/buffers.ts | Adds circular buffers + shared logging entry types |
| .claude/skills/gstack/browse/scripts/build-node-server.sh | Adds Node-compatible server bundling script for Windows |
| .claude/skills/gstack/browse/SKILL.md.tmpl | Adds /browse skill template docs for commands and workflows |
| .claude/skills/gstack/benchmark/SKILL.md.tmpl | Adds /benchmark performance regression workflow template |
| .claude/skills/gstack/agents/openai.yaml | Adds agent interface metadata for gstack bundle |
| .claude/skills/gstack/actionlint.yaml | Adds actionlint config (runner label) |
| .claude/skills/gstack/VERSION | Adds gstack version file |
| .claude/skills/gstack/SKILL.md.tmpl | Adds top-level gstack skill template |
| .claude/skills/gstack/README.md | Adds gstack README (installation, workflows, tooling overview) |
| .claude/skills/gstack/LICENSE | Adds MIT license |
| .claude/skills/gstack/ETHOS.md | Adds builder ethos documentation |
| .claude/skills/gstack/CLAUDE.md | Adds contributor/dev workflow documentation |
| .claude/skills/gstack/BROWSER.md | Adds browser internals + command reference documentation |
| .claude/skills/gstack/AGENTS.md | Adds skills overview for non-Claude hosts |
| .claude/skills/gstack/.gitignore | Adds gstack-local ignore rules (dist, logs, state dirs, etc.) |
| .claude/skills/gstack/.github/workflows/skill-docs.yml | Adds CI check to enforce generated skill docs freshness |
| .claude/skills/gstack/.github/workflows/evals.yml | Adds CI E2E eval workflow with docker caching + PR reporting |
| .claude/skills/gstack/.github/workflows/ci-image.yml | Adds workflow to build/push CI runner image |
| .claude/skills/gstack/.github/workflows/actionlint.yml | Adds workflow linting job via actionlint |
| .claude/skills/gstack/.github/docker/Dockerfile.ci | Adds CI docker image definition with Bun/Node/Playwright/Claude CLI |
| .claude/skills/gstack/.github/actionlint.yaml | Adds self-hosted runner label config for actionlint |
| .claude/skills/gstack/.env.example | Adds example env file for evals |
| --- | ||
| name: careful | ||
| version: 0.1.0 | ||
| description: | | ||
| MANUAL TRIGGER ONLY: invoke only when user types /careful. |
There was a problem hiding this comment.
The PR title/description is about a "bear race mechanic", but the actual changes vendor the gstack skill suite and implement/extend a Playwright-based browse daemon (+ skills/docs/CI). Please update either the PR title/description to reflect the gstack/browse work, or split these changes into the correct PR(s).
| case 'upload': { | ||
| const [selector, ...filePaths] = args; | ||
| if (!selector || filePaths.length === 0) throw new Error('Usage: browse upload <selector> <file1> [file2...]'); | ||
|
|
||
| // Validate all files exist before upload | ||
| for (const fp of filePaths) { | ||
| if (!fs.existsSync(fp)) throw new Error(`File not found: ${fp}`); | ||
| } |
There was a problem hiding this comment.
The new upload command can read arbitrary local files and upload them to any page, which is a high-risk exfiltration vector for an agent-capable tool. Consider restricting uploads to the same safe directories used elsewhere (TEMP_DIR, process.cwd()), and reject traversal (e.g., ../) similarly to validateReadPath/validateOutputPath.
| const normalized = path.normalize(filePath); | ||
| if (normalized.includes('..')) { | ||
| throw new Error('Path traversal sequences (..) are not allowed'); | ||
| } |
There was a problem hiding this comment.
This traversal check will also block legitimate paths that merely contain .. in a filename (e.g., notes..backup.js). To avoid false positives (and also handle both separators), check for .. as a path segment (e.g., split by path.sep after normalization, or use a regex like /(^|[\\\\/])\\.\\.([\\\\/]|$)/). The same pattern appears elsewhere (e.g., cookie import).
|
|
||
| /** Check if resolvedPath is within dir, using platform-aware separators. */ | ||
| export function isPathWithin(resolvedPath: string, dir: string): boolean { | ||
| return resolvedPath === dir || resolvedPath.startsWith(dir + path.sep); |
There was a problem hiding this comment.
On Windows, path comparisons are case-insensitive and may contain mixed separators, so this string-prefix check can incorrectly reject valid paths (or behave inconsistently) if casing differs (e.g., C:\\Temp vs c:\\temp). A more robust approach is to normalize/resolve both sides (and on Windows, compare with a consistent casing) before checking containment.
| return resolvedPath === dir || resolvedPath.startsWith(dir + path.sep); | |
| const normalizedPath = path.resolve(resolvedPath); | |
| const normalizedDir = path.resolve(dir); | |
| const pathToCompare = IS_WINDOWS ? normalizedPath.toLowerCase() : normalizedPath; | |
| const dirToCompare = IS_WINDOWS ? normalizedDir.toLowerCase() : normalizedDir; | |
| const dirWithSep = dirToCompare.endsWith(path.sep) | |
| ? dirToCompare | |
| : dirToCompare + path.sep; | |
| return pathToCompare === dirToCompare || pathToCompare.startsWith(dirWithSep); |
| head -1 "$DIST_DIR/server-node.mjs" | ||
| echo '// ── Windows Node.js compatibility (auto-generated) ──' | ||
| echo 'import { fileURLToPath as _ftp } from "node:url";' | ||
| echo 'import { dirname as _dn } from "node:path";' |
There was a problem hiding this comment.
The injected header uses createRequire(...) but doesn’t import it, which will throw at runtime in Node ESM (ReferenceError: createRequire is not defined). Import it explicitly (e.g., from node:module) in the injected header so the generated server-node.mjs is self-contained.
| echo 'import { dirname as _dn } from "node:path";' | |
| echo 'import { dirname as _dn } from "node:path";' | |
| echo 'import { createRequire } from "node:module";' |
| const screenshotPath = opts.outputPath || `${TEMP_DIR}/browse-annotated.png`; | ||
| // Validate output path (consistent with screenshot/pdf/responsive) | ||
| const resolvedPath = require('path').resolve(screenshotPath); | ||
| const safeDirs = [TEMP_DIR, process.cwd()]; | ||
| if (!safeDirs.some((dir: string) => isPathWithin(resolvedPath, dir))) { | ||
| throw new Error(`Path must be within: ${safeDirs.join(', ')}`); | ||
| } |
There was a problem hiding this comment.
snapshot.ts is TypeScript/ESM, but this block uses require('path'). That will break when running in a pure ESM Node bundle (and is also inconsistent with other modules that import * as path from 'path'). Prefer a static import at the top and use path.resolve(...) here.
| afterAll(() => { | ||
| try { testServer.server.stop(); } catch {} | ||
| setTimeout(() => process.exit(0), 500); |
There was a problem hiding this comment.
Calling process.exit(0) from afterAll will terminate the entire Bun test process and can prevent other test files from running/finishing when the full suite is executed. Replace this with explicit resource cleanup (await bm.close(), ensure all Playwright instances are closed, stop the server), and let the test runner exit naturally.
| afterAll(() => { | |
| try { testServer.server.stop(); } catch {} | |
| setTimeout(() => process.exit(0), 500); | |
| afterAll(async () => { | |
| try { | |
| if (bm) { | |
| await bm.close(); | |
| } | |
| } catch {} | |
| try { | |
| testServer.server.stop(); | |
| } catch {} |
| test('priority chain checks .codex, .agents, .claude markers', () => { | ||
| // Verify the source code implements the correct priority order. | ||
| // We read the function source to confirm the markers array order. | ||
| const src = require('fs').readFileSync(require('path').join(__dirname, '../src/find-browse.ts'), 'utf-8'); |
There was a problem hiding this comment.
This test mixes CommonJS globals (require, __dirname) inside an ESM-style test file (import { describe, test, expect } ...). In many Bun/TS ESM contexts, __dirname (and sometimes require) is undefined, causing a runtime crash. Use import.meta.dir for the directory and regular import (or await import(...)) patterns instead.
Summary of ChangesHello, I'm Gemini Code Assist1! I'm currently reviewing this pull request and will post my feedback shortly. In the meantime, here's a summary to help you and other reviewers quickly get up to speed! This pull request introduces a comprehensive bear racing system, transforming passive mountain bears into active competitors. The new mechanics aim to enhance player engagement by providing dynamic challenges and clear feedback on race outcomes, alongside crucial fixes to game mode transitions. Highlights
Using Gemini Code AssistThe full guide for Gemini Code Assist can be found on our documentation page, here are some quick tips. Invoking Gemini You can request assistance from Gemini at any point by creating a comment using either
Customization To customize Gemini Code Assist for GitHub experience, repository maintainers can create a configuration file and/or provide a custom code review style guide (such as PEP-8 for Python) by creating and adding files to a Limitations & Feedback Gemini Code Assist may make mistakes. Please leave feedback on any instances where its feedback is incorrect or counter productive. You can react with 👍 and 👎 on @gemini-code-assist comments. If you're interested in giving your feedback about your experience with Gemini Code Assist for GitHub and other Google products, sign up here. Footnotes
|
![gemini-code-assist[bot]](https://avatars.githubusercontent.com/in/956858?s=60&v=4){kind=small}
There was a problem hiding this comment.
Code Review
This pull request introduces significant infrastructure and documentation updates for the gstack project. Key changes include new GitHub Actions for workflow linting, CI image building, and E2E evaluations, alongside a refined Dockerfile for the CI environment. Extensive documentation has been added or updated, covering agent skills, architecture, browser reference, development guidelines, and project ethos. New utility scripts for development setup, telemetry, and git repository analysis have also been introduced or modified. Review feedback highlights several areas for improvement, including addressing an insecure chmod command in the Dockerfile, consolidating apt-get update commands for efficiency, improving shell script portability, removing a duplicated .env entry, correcting telemetry processing logic that prematurely breaks loops, using jq for robust JSON parsing, and eliminating redundant slug generation scripts.
| && mkdir -p /home/runner/.gstack && chown -R runner:runner /home/runner/.gstack \ | ||
| && chmod 1777 /tmp \ | ||
| && mkdir -p /home/runner/.bun && chown -R runner:runner /home/runner/.bun \ | ||
| && chmod -R 1777 /tmp |
There was a problem hiding this comment.
The command chmod -R 1777 /tmp is insecure as it makes all files and subdirectories in /tmp world-writable, which can lead to security vulnerabilities. The correct permission for the /tmp directory itself is 1777 (with the sticky bit), which is already correctly set on line 61. This recursive chmod is unnecessary and should be removed to improve security.
| RUN apt-get update && apt-get install -y --no-install-recommends \ | ||
| git curl unzip ca-certificates jq bc gpg \ | ||
| && rm -rf /var/lib/apt/lists/* | ||
|
|
||
| # GitHub CLI | ||
| RUN curl -fsSL https://cli.github.com/packages/githubcli-archive-keyring.gpg \ | ||
| | gpg --dearmor -o /usr/share/keyrings/githubcli-archive-keyring.gpg \ | ||
| && echo "deb [arch=$(dpkg --print-architecture) signed-by=/usr/share/keyrings/githubcli-archive-keyring.gpg] https://cli.github.com/packages stable main" \ | ||
| | tee /etc/apt/sources.list.d/github-cli.list > /dev/null \ | ||
| && apt-get update && apt-get install -y --no-install-recommends gh \ | ||
| && rm -rf /var/lib/apt/lists/* | ||
|
|
||
| # Node.js 22 LTS (needed for claude CLI) | ||
| RUN curl -fsSL https://deb.nodesource.com/setup_22.x | bash - \ | ||
| && apt-get install -y --no-install-recommends nodejs \ | ||
| && rm -rf /var/lib/apt/lists/* |
There was a problem hiding this comment.
The Dockerfile has multiple RUN apt-get update && ... && rm -rf /var/lib/apt/lists/* commands. This is inefficient as it creates multiple layers and runs apt-get update repeatedly. It's a best practice to combine these into a single RUN instruction to reduce image size and build time. Consider performing all package source setup first, then a single apt-get update, followed by a single apt-get install for all packages, and finally a single cleanup step.
|
|
||
| | Suite | Result | Status | Cost | | ||
| |-------|--------|--------|------| | ||
| $(echo -e "$SUITE_LINES") |
There was a problem hiding this comment.
Using echo -e can have inconsistent behavior across different shells and systems. For more portable and safer interpretation of backslash escapes like \n, it's recommended to use printf. This will ensure the markdown table for the PR comment is always formatted correctly.
$(printf '%b' "$SUITE_LINES")| *.log | ||
| bun.lock | ||
| *.bun-build | ||
| .env |
There was a problem hiding this comment.
The entry .env is duplicated. It's already present on line 1. You can remove this redundant entry.
| mkdir -p ~/.gstack/analytics | ||
| echo '{"skill":"gstack","ts":"'$(date -u +%Y-%m-%dT%H:%M:%SZ)'","repo":"'$(basename "$(git rev-parse --show-toplevel 2>/dev/null)" 2>/dev/null || echo "unknown")'"}' >> ~/.gstack/analytics/skill-usage.jsonl 2>/dev/null || true | ||
| # zsh-compatible: use find instead of glob to avoid NOMATCH error | ||
| for _PF in $(find ~/.gstack/analytics -maxdepth 1 -name '.pending-*' 2>/dev/null); do [ -f "$_PF" ] && ~/.claude/skills/gstack/bin/gstack-telemetry-log --event-type skill_run --skill _pending_finalize --outcome unknown --session-id "$_SESSION_ID" 2>/dev/null || true; break; done |
There was a problem hiding this comment.
The loop to finalize stale pending telemetry events has a break statement, which means it will only process one stale file per run, even if multiple exist. This could lead to a backlog of unprocessed pending events. To ensure all stale files are processed, the break should be removed.
| for _PF in $(find ~/.gstack/analytics -maxdepth 1 -name '.pending-*' 2>/dev/null); do [ -f "$_PF" ] && ~/.claude/skills/gstack/bin/gstack-telemetry-log --event-type skill_run --skill _pending_finalize --outcome unknown --session-id "$_SESSION_ID" 2>/dev/null || true; break; done | |
| for _PF in $(find ~/.gstack/analytics -maxdepth 1 -name '.pending-*' 2>/dev/null); do [ -f "$_PF" ] && ~/.claude/skills/gstack/bin/gstack-telemetry-log --event-type skill_run --skill _pending_finalize --outcome unknown --session-id "$_SESSION_ID" 2>/dev/null || true; done |
| WEEKLY="$(echo "$DATA" | grep -o '"weekly_active":[0-9]*' | grep -o '[0-9]*' || echo "0")" | ||
| CHANGE="$(echo "$DATA" | grep -o '"change_pct":[0-9-]*' | grep -o '[0-9-]*' || echo "0")" |
There was a problem hiding this comment.
Parsing JSON with grep and awk is fragile and can break if the JSON formatting changes (e.g., with extra whitespace). Since jq is installed as part of the environment (as seen in the Dockerfile), it would be more robust to use jq for parsing the JSON response from the Supabase edge function. This applies to other JSON parsing in this script as well.
WEEKLY="$(echo "$DATA" | jq -r '.weekly_active // 0')"
CHANGE="$(echo "$DATA" | jq -r '.change_pct // 0')"
| REMOTE_URL=$(git remote get-url origin 2>/dev/null || true) | ||
| if [ -z "$REMOTE_URL" ]; then | ||
| echo "REPO_MODE=unknown" | ||
| exit 0 | ||
| fi | ||
| SLUG=$(echo "$REMOTE_URL" | sed 's|.*[:/]\([^/]*/[^/]*\)\.git$|\1|;s|.*[:/]\([^/]*/[^/]*\)$|\1|' | tr '/' '-') | ||
| [ -z "${SLUG:-}" ] && { echo "REPO_MODE=unknown"; exit 0; } |
There was a problem hiding this comment.
This script re-implements the logic for getting the repository slug, which is also present in the gstack-slug script. To avoid duplication and ensure consistency (e.g., sanitization of the slug), you could call gstack-slug from this script. The main gstack-slug script also includes sanitization which makes it more robust.
eval "$($SCRIPT_DIR/gstack-slug 2>/dev/null)"
[ -z "${SLUG:-}" ] && { echo "REPO_MODE=unknown"; exit 0; }
| If `PROACTIVE` is `"false"`, do not proactively suggest gstack skills — only invoke | ||
| them when the user explicitly asks. The user opted out of proactive suggestions. | ||
|
|
||
| If output shows `UPGRADE_AVAILABLE <old> <new>`: read `~/.claude/skills/gstack/gstack-upgrade/SKILL.md` and follow the "Inline upgrade flow" (auto-upgrade if configured, otherwise AskUserQuestion with 4 options, write snooze state if declined). If `JUST_UPGRADED <from> <to>`: tell user "Running gstack v{to} (just updated!)" and continue. |
There was a problem hiding this comment.
The loop to finalize stale pending telemetry events has a break statement, which means it will only process one stale file per run, even if multiple exist. This could lead to a backlog of unprocessed pending events. To ensure all stale files are processed, the break should be removed.
| If output shows `UPGRADE_AVAILABLE <old> <new>`: read `~/.claude/skills/gstack/gstack-upgrade/SKILL.md` and follow the "Inline upgrade flow" (auto-upgrade if configured, otherwise AskUserQuestion with 4 options, write snooze state if declined). If `JUST_UPGRADED <from> <to>`: tell user "Running gstack v{to} (just updated!)" and continue. | |
| for _PF in $(find ~/.gstack/analytics -maxdepth 1 -name '.pending-*' 2>/dev/null); do [ -f "$_PF" ] && ~/.claude/skills/gstack/bin/gstack-telemetry-log --event-type skill_run --skill _pending_finalize --outcome unknown --session-id "$_SESSION_ID" 2>/dev/null || true; done |
| #!/usr/bin/env bash | ||
| # Output the remote slug (owner-repo) for the current git repo. | ||
| # Used by SKILL.md files to derive project-specific paths in ~/.gstack/projects/. | ||
| set -e | ||
| URL=$(git remote get-url origin 2>/dev/null || true) | ||
| if [ -n "$URL" ]; then | ||
| # Strip trailing .git if present, then extract owner/repo | ||
| URL="${URL%.git}" | ||
| # Handle both SSH (git@host:owner/repo) and HTTPS (https://host/owner/repo) | ||
| OWNER_REPO=$(echo "$URL" | sed -E 's#.*[:/]([^/]+)/([^/]+)$#\1-\2#') | ||
| echo "$OWNER_REPO" | ||
| else | ||
| basename "$(git rev-parse --show-toplevel 2>/dev/null || pwd)" | ||
| fi |
There was a problem hiding this comment.
This script duplicates functionality found in bin/gstack-slug. To reduce code duplication and ensure consistency, consider removing this script and using the more robust bin/gstack-slug script, which also includes output sanitization.
| mkdir -p ~/.gstack/analytics | ||
| echo '{"skill":"canary","ts":"'$(date -u +%Y-%m-%dT%H:%M:%SZ)'","repo":"'$(basename "$(git rev-parse --show-toplevel 2>/dev/null)" 2>/dev/null || echo "unknown")'"}' >> ~/.gstack/analytics/skill-usage.jsonl 2>/dev/null || true | ||
| # zsh-compatible: use find instead of glob to avoid NOMATCH error | ||
| for _PF in $(find ~/.gstack/analytics -maxdepth 1 -name '.pending-*' 2>/dev/null); do [ -f "$_PF" ] && ~/.claude/skills/gstack/bin/gstack-telemetry-log --event-type skill_run --skill _pending_finalize --outcome unknown --session-id "$_SESSION_ID" 2>/dev/null || true; break; done |
There was a problem hiding this comment.
The loop to finalize stale pending telemetry events has a break statement, which means it will only process one stale file per run, even if multiple exist. This could lead to a backlog of unprocessed pending events. To ensure all stale files are processed, the break should be removed.
| for _PF in $(find ~/.gstack/analytics -maxdepth 1 -name '.pending-*' 2>/dev/null); do [ -f "$_PF" ] && ~/.claude/skills/gstack/bin/gstack-telemetry-log --event-type skill_run --skill _pending_finalize --outcome unknown --session-id "$_SESSION_ID" 2>/dev/null || true; break; done | |
| for _PF in $(find ~/.gstack/analytics -maxdepth 1 -name '.pending-*' 2>/dev/null); do [ -f "$_PF" ] && ~/.claude/skills/gstack/bin/gstack-telemetry-log --event-type skill_run --skill _pending_finalize --outcome unknown --session-id "$_SESSION_ID" 2>/dev/null || true; done |
Bears that were already living on the mountain now become race opponents the moment the player starts
sledding downhill. They steer toward the player to jostle and bump them (bumpercars-style), creating a
natural cascade of competition as the player descends past more bears.
What's in this PR:
applying a bump impulse on close contact
the race each frame as the player descends; no one-time trigger
close
visible screen so the player always has competition
(win) / gold (mixed) / brown (loss)
DOWNHILL; Tab still works anywhere as a manual toggle