Skip to content

Tayhlow8/Caixa-D-gua-website-2026 / .agents/skills/gstack

gstack

Flagged

Fast headless browser for QA testing and site dogfooding. Navigate pages, interact with elements, verify state, diff before/after, take annotated screenshots, test responsive layouts, forms, uploads, dialogs, and capture bug evidence. Use when asked to open or test a site, verify a deployment, dogfood a user flow, or file a bug with screenshots. (gstack)

FileKindSize
AGENTS.mdmarkdown2.5 KB
ARCHITECTURE.mdmarkdown29.3 KB
BROWSER.mdmarkdown29.2 KB
CHANGELOG.mdmarkdown389.3 KB
CLAUDE.mdmarkdown40.4 KB
CONTRIBUTING.mdmarkdown21.4 KB
DESIGN.mdmarkdown4.4 KB
ETHOS.mdmarkdown7.3 KB
LICENSEother1.0 KB
README.mdmarkdown37.3 KB
SKILL.mdskill md40.7 KB
SKILL.md.tmplother10.9 KB
TODOS.mdmarkdown79.0 KB
USING_GBRAIN_WITH_GSTACK.mdmarkdown19.0 KB
VERSIONother9 B
actionlint.yamlconfig56 B
agents/openai.yamlconfig351 B
bin/chrome-cdpother2.2 KB
bin/dev-setupother2.4 KB
bin/dev-teardownother1.6 KB
bin/gstack-analyticsother6.3 KB
bin/gstack-brain-consumerother5.5 KB
bin/gstack-brain-enqueueother1.9 KB
bin/gstack-brain-initother13.8 KB
bin/gstack-brain-readerother21 B
bin/gstack-brain-restoreother7.5 KB
bin/gstack-brain-syncother14.6 KB
bin/gstack-brain-uninstallother5.2 KB
bin/gstack-builder-profileother616 B
bin/gstack-codex-probeother4.5 KB
bin/gstack-community-dashboardother4.5 KB
bin/gstack-configother10.4 KB
bin/gstack-developer-profileother15.9 KB
bin/gstack-diff-scopeother3.1 KB
bin/gstack-extensionother2.0 KB
bin/gstack-gbrain-detectother3.6 KB
bin/gstack-gbrain-installother6.6 KB
bin/gstack-gbrain-lib.shscript3.4 KB
bin/gstack-gbrain-repo-policyother7.0 KB
bin/gstack-gbrain-supabase-provisionother15.2 KB
bin/gstack-gbrain-supabase-verifyother4.0 KB
bin/gstack-global-discover.tsscript18.8 KB
bin/gstack-jsonl-mergeother2.7 KB
bin/gstack-learnings-logother3.3 KB
bin/gstack-learnings-searchother4.8 KB
bin/gstack-model-benchmarkother5.8 KB
bin/gstack-next-versionother16.3 KB
bin/gstack-open-urlother338 B
bin/gstack-patch-namesother1.2 KB
bin/gstack-platform-detectother1.0 KB
bin/gstack-question-logother6.4 KB
bin/gstack-question-preferenceother9.4 KB
bin/gstack-relinkother3.0 KB
bin/gstack-repo-modeother3.5 KB
bin/gstack-review-logother866 B
bin/gstack-review-readother528 B
bin/gstack-security-dashboardother4.9 KB
bin/gstack-session-updateother3.5 KB
bin/gstack-settings-hookother2.8 KB
bin/gstack-slugother1.9 KB
bin/gstack-specialist-statsother2.2 KB
bin/gstack-taste-updateother10.5 KB
bin/gstack-team-initother6.7 KB
bin/gstack-telemetry-logother10.0 KB
bin/gstack-telemetry-syncother5.3 KB
bin/gstack-timeline-logother1.6 KB
bin/gstack-timeline-readother2.6 KB
bin/gstack-uninstallother10.3 KB
bin/gstack-update-checkother7.7 KB
bun.lockother57.4 KB
claude/SKILL.md.tmplother8.7 KB
conductor.jsonconfig87 B
connect-chromeother19 B
contrib/add-host/SKILL.md.tmplother1.7 KB
design/prototype.tsscript5.4 KB
design/src/auth.tsscript1.7 KB
design/src/brief.tsscript1.9 KB
design/src/check.tsscript3.3 KB
design/src/cli.tsscript8.9 KB
design/src/commands.tsscript3.0 KB
design/src/compare.tsscript19.1 KB
design/src/design-to-code.tsscript3.0 KB
design/src/diff.tsscript3.3 KB
design/src/evolve.tsscript5.3 KB
design/src/gallery.tsscript7.0 KB
design/src/generate.tsscript4.4 KB
design/src/iterate.tsscript6.5 KB
design/src/memory.tsscript6.0 KB
design/src/serve.tsscript8.3 KB
design/src/session.tsscript1.8 KB
design/src/variants.tsscript8.7 KB
design/test/feedback-roundtrip.test.tsscript12.6 KB
design/test/gallery.test.tsscript5.0 KB
design/test/serve.test.tsscript16.0 KB
docs/ADDING_A_HOST.mdmarkdown6.1 KB
docs/ON_THE_LOC_CONTROVERSY.mdmarkdown13.2 KB
docs/OPENCLAW.mdmarkdown6.5 KB
docs/REMOTE_BROWSER_ACCESS.mdmarkdown9.6 KB
docs/designs/BUN_NATIVE_INFERENCE.mdmarkdown6.7 KB
docs/designs/CHROME_VS_CHROMIUM_EXPLORATION.mdmarkdown4.0 KB
docs/designs/CONDUCTOR_CHROME_SIDEBAR_INTEGRATION.mdmarkdown3.5 KB
docs/designs/CONDUCTOR_SESSION_API.mdmarkdown3.9 KB
docs/designs/DESIGN_SHOTGUN.mdmarkdown22.5 KB
docs/designs/DESIGN_TOOLS_V1.mdmarkdown36.0 KB
docs/designs/GCOMPACTION.mdmarkdown57.6 KB
docs/designs/GSTACK_BROWSER_V0.mdmarkdown17.3 KB
docs/designs/ML_PROMPT_INJECTION_KILLER.mdmarkdown20.7 KB
docs/designs/PACING_UPDATES_V0.mdmarkdown9.7 KB
docs/designs/PLAN_TUNING_V0.mdmarkdown31.2 KB
docs/designs/PLAN_TUNING_V1.mdmarkdown19.0 KB
docs/designs/SELF_LEARNING_V0.mdmarkdown15.4 KB
docs/designs/SESSION_INTELLIGENCE.mdmarkdown6.8 KB
docs/designs/SIDEBAR_MESSAGE_FLOW.mdmarkdown10.1 KB
docs/designs/SLATE_HOST.mdmarkdown12.6 KB
docs/designs/SLOP_SCAN_FOR_REVIEW_SHIP.mdmarkdown3.0 KB
docs/evals/security-bench-ensemble-v2.jsonconfig1.6 KB
docs/gbrain-sync-errors.mdmarkdown6.2 KB
docs/gbrain-sync.mdmarkdown6.5 KB
docs/images/github-2013.pngbinary61.9 KB
docs/images/github-2026.pngbinary59.1 KB
docs/skills.mdmarkdown59.5 KB
extension/background.jsscript20.6 KB
extension/content.cssother2.5 KB
extension/content.jsscript11.9 KB
extension/icons/icon-128.pngbinary2.8 KB
extension/icons/icon-16.pngbinary400 B
extension/icons/icon-48.pngbinary1.1 KB
extension/inspector.cssother798 B
extension/inspector.jsscript15.7 KB
extension/lib/xterm-addon-fit.jsscript1.5 KB
extension/lib/xterm.cssother5.3 KB
extension/lib/xterm.jsscript276.8 KB
extension/manifest.jsonconfig783 B
extension/popup.htmlother2.3 KB
extension/popup.jsscript1.7 KB
extension/sidepanel-terminal.jsscript15.2 KB
extension/sidepanel.cssother39.6 KB
extension/sidepanel.htmlother9.1 KB
extension/sidepanel.jsscript40.2 KB
hosts/claude.tsscript1.1 KB
hosts/codex.tsscript2.2 KB
hosts/cursor.tsscript1.1 KB
hosts/factory.tsscript1.8 KB
hosts/gbrain.tsscript2.2 KB
hosts/hermes.tsscript2.1 KB
hosts/index.tsscript2.2 KB
hosts/kiro.tsscript1.2 KB
hosts/openclaw.tsscript2.0 KB
hosts/opencode.tsscript1.3 KB
hosts/slate.tsscript1.1 KB
lib/worktree.tsscript9.6 KB
model-overlays/claude.mdmarkdown613 B
model-overlays/gemini.mdmarkdown551 B
model-overlays/gpt-5.4.mdmarkdown683 B
model-overlays/gpt.mdmarkdown1.9 KB
model-overlays/o-series.mdmarkdown660 B
model-overlays/opus-4-7.mdmarkdown1.5 KB
openclaw/agents-gstack-section.mdmarkdown2.5 KB
openclaw/gstack-full-CLAUDE.mdmarkdown538 B
openclaw/gstack-lite-CLAUDE.mdmarkdown688 B
openclaw/gstack-plan-CLAUDE.mdmarkdown1.0 KB
package.jsonmanifest4.7 KB
scripts/analytics.tsscript5.3 KB
scripts/app/gstack-browserother2.4 KB
scripts/app/icon.icnsother1.0 MB
scripts/archetypes.tsscript5.4 KB
scripts/build-app.shscript7.5 KB
scripts/compare-pr-version.tsscript3.0 KB
scripts/detect-bump.tsscript1.2 KB
scripts/dev-skill.tsscript2.4 KB
scripts/discover-skills.tsscript1.2 KB
scripts/eval-compare.tsscript3.0 KB
scripts/eval-list.tsscript3.5 KB
scripts/eval-select.tsscript3.1 KB
scripts/eval-summary.tsscript6.5 KB
scripts/eval-watch.tsscript5.4 KB
scripts/garry-output-comparison.tsscript17.0 KB
scripts/gen-skill-docs.tsscript28.3 KB
scripts/host-adapters/openclaw-adapter.tsscript1.7 KB
scripts/host-config-export.tsscript3.4 KB
scripts/host-config.tsscript7.3 KB
scripts/jargon-list.jsonconfig1.8 KB
scripts/models.tsscript2.5 KB
scripts/one-way-doors.tsscript5.6 KB
scripts/preflight-agent-sdk.tsscript4.4 KB
scripts/psychographic-signals.tsscript10.3 KB
scripts/question-registry.tsscript23.5 KB
scripts/resolvers/browse.tsscript6.2 KB
scripts/resolvers/codex-helpers.tsscript5.0 KB
scripts/resolvers/composition.tsscript1.8 KB
scripts/resolvers/confidence.tsscript1.6 KB
scripts/resolvers/constants.tsscript3.1 KB
scripts/resolvers/design.tsscript56.9 KB
scripts/resolvers/dx.tsscript5.4 KB
scripts/resolvers/gbrain.tsscript5.3 KB
scripts/resolvers/index.tsscript4.3 KB
scripts/resolvers/learnings.tsscript3.9 KB
scripts/resolvers/make-pdf.tsscript2.2 KB
scripts/resolvers/model-overlay.tsscript2.2 KB
scripts/resolvers/preamble.tsscript5.7 KB

scripts/one-way-doors.ts

View on GitHub
/**
 * One-Way Door Classifier — belt-and-suspenders safety layer.
 *
 * Primary safety gate is the `door_type` field in scripts/question-registry.ts.
 * Every registered AskUserQuestion declares whether it is one-way (always ask,
 * never auto-decide) or two-way (can be suppressed by explicit user preference).
 *
 * This file is a SECONDARY keyword-pattern check for questions that fire
 * WITHOUT a registry id (ad-hoc question_ids generated at runtime). If the
 * question_summary contains any of the destructive keyword patterns, treat
 * it as one-way regardless of what the (absent or unknown) registry entry says.
 *
 * Codex correctly pointed out (design doc Decision C) that prose-parsing is
 * too weak to be the PRIMARY safety gate — wording can change. The registry
 * is primary. This is the fallback for questions not yet catalogued, and it
 * errs on the side of asking the user even when tuning preferences say skip.
 *
 * Ordering
 * --------
 * isOneWayDoor() is called by gstack-question-sensitivity --check in this
 * order:
 *   1. Look up registry by id → use registry.door_type if found
 *   2. If not in registry: apply keyword patterns below
 *   3. Default to ASK_NORMALLY (safer than AUTO_DECIDE)
 */

import { getQuestion } from './question-registry';

/**
 * Keyword patterns that identify one-way-door questions when the registry
 * doesn't have an entry for the question_id. Case-insensitive substring match
 * against the question_summary passed into AskUserQuestion.
 *
 * Additions here should be conservative — a false positive means the user
 * gets asked an extra question they might have preferred to auto-decide.
 * A false negative could mean auto-approving a destructive operation.
 */
const DESTRUCTIVE_PATTERNS: RegExp[] = [
  // File system destruction
  /\brm\s+-rf\b/i,
  /\bdelete\b/i,
  /\bremove\s+(directory|folder|files?)\b/i,
  /\bwipe\b/i,
  /\bpurge\b/i,
  /\btruncate\b/i,

  // Database destruction
  /\bdrop\s+(table|database|schema|index|column)\b/i,
  /\bdelete\s+from\b/i,

  // Git / VCS destruction
  /\bforce[- ]push\b/i,
  /\bpush\s+--force\b/i,
  /\bgit\s+reset\s+--hard\b/i,
  /\bcheckout\s+--\b/i,
  /\brestore\s+\.\b/i,
  /\bclean\s+-f\b/i,
  /\bbranch\s+-D\b/i,

  // Deploy / infra destruction
  /\bkubectl\s+delete\b/i,
  /\bterraform\s+destroy\b/i,
  /\brollback\b/i,

  // Credentials / auth — allow filler words ("the", "my") between verb and noun
  /\brevoke\s+[\w\s]*\b(api key|token|credential|access key|password)\b/i,
  /\breset\s+[\w\s]*\b(api key|token|password|credential)\b/i,
  /\brotate\s+[\w\s]*\b(api key|token|secret|credential|access key)\b/i,

  // Scope / architecture forks (reversible with effort — still deserve confirmation)
  /\barchitectur(e|al)\s+(change|fork|shift|decision)\b/i,
  /\bdata\s+model\s+change\b/i,
  /\bschema\s+migration\b/i,
  /\bbreaking\s+change\b/i,
];

/**
 * Skill-category combinations that are always one-way even when the question
 * body looks benign. Matches the ownership model: certain skill actions are
 * inherently high-stakes.
 */
const ONE_WAY_SKILL_CATEGORIES = new Set<string>([
  'cso:approval', // security-audit findings
  'land-and-deploy:approval', // anything /land-and-deploy asks
]);

export interface ClassifyInput {
  /** Registry id OR ad-hoc id; looked up first */
  question_id?: string;
  /** Skill firing the question (for skill-category fallback) */
  skill?: string;
  /** Question category (approval | clarification | routing | cherry-pick | feedback-loop) */
  category?: string;
  /** Free-form question summary — pattern-matched against destructive keywords */
  summary?: string;
}

export interface ClassifyResult {
  /** true = treat as one-way door (always ask, never auto-decide) */
  oneWay: boolean;
  /** Which check triggered the classification (for audit/debug) */
  reason: 'registry' | 'skill-category' | 'keyword' | 'default-safe' | 'default-two-way';
  /** Matched pattern if reason is 'keyword' */
  matched?: string;
}

/**
 * Classify a question as one-way (always ask) or two-way (can be suppressed).
 * Returns {oneWay: false, reason: 'default-two-way'} only when no evidence of
 * one-way nature is found. Errs conservatively otherwise.
 */
export function classifyQuestion(input: ClassifyInput): ClassifyResult {
  // 1. Registry lookup (primary)
  if (input.question_id) {
    const registered = getQuestion(input.question_id);
    if (registered) {
      return {
        oneWay: registered.door_type === 'one-way',
        reason: 'registry',
      };
    }
  }

  // 2. Skill-category fallback (certain combos are always one-way)
  if (input.skill && input.category) {
    const key = `${input.skill}:${input.category}`;
    if (ONE_WAY_SKILL_CATEGORIES.has(key)) {
      return { oneWay: true, reason: 'skill-category' };
    }
  }

  // 3. Keyword pattern match (catch destructive questions without registry entry)
  if (input.summary) {
    for (const pattern of DESTRUCTIVE_PATTERNS) {
      if (pattern.test(input.summary)) {
        return {
          oneWay: true,
          reason: 'keyword',
          matched: pattern.toString(),
        };
      }
    }
  }

  // 4. No evidence either way — treat as two-way (can be preference-suppressed).
  return { oneWay: false, reason: 'default-two-way' };
}

/**
 * Convenience wrapper for the sensitivity check binary.
 * Returns true if the question must be asked regardless of user preferences.
 */
export function isOneWayDoor(input: ClassifyInput): boolean {
  return classifyQuestion(input).oneWay;
}

/**
 * Export patterns for tests and audit tooling.
 */
export const DESTRUCTIVE_PATTERN_LIST = DESTRUCTIVE_PATTERNS;
export const ONE_WAY_SKILL_CATEGORY_SET = ONE_WAY_SKILL_CATEGORIES;
finishing-a-development-branchobra/superpowersUse when implementation is complete, all tests pass, and you need to decide how to integrate the workStars: 322kUpdated Claude CodeCodexVerifiedInstall
systematic-debuggingobra/superpowersUse when encountering any bug, test failure, or unexpected behavior, before proposing fixesStars: 322kUpdated Claude CodeCodexVerifiedInstall
test-driven-developmentobra/superpowersUse when implementing any feature or bugfix, before writing implementation codeStars: 322kUpdated Claude CodeCodexVerifiedInstall
skill-creatoranthropics/skillsCreate new skills, modify and improve existing skills, and measure skill performance. Use when users want to create a skill from scratch, edit, or optimize an existing skill, run evals to test a skill, benchmark skill performance with variance analysis, or optimize a skill's description for better triggering accuracy.Stars: 198kUpdated Claude CodeCodexVerifiedInstall