Searches and enumerates specialized skills from the offline vault (915+ skills) by keyword or category. Safe, read-only operation: reads local catalog.json with zero side effects, no mutations, and no network access. When query and category are both provided, applies a logical AND filter. When both are omitted or empty, returns the top general engineering skills up to the specified limit. If no skills match, returns a helpful message with suggested search keywords.

When to use: Call to discover skill names and capabilities before installation. When NOT to use: Do not use to install or execute skills; call equip_skill once a specific skill name is chosen. Returns: Formatted list of matching skills containing name, category tag, and summary description, plus remaining match count.

Extracts and installs a specialized skill package from the offline vault into the assistant's active skills folder. Mutating file operation: creates or overwrites only the specific skill subfolder in the target directory (defaulting to ~/.gemini/config/skills); leaves all other workspace files untouched. Requires no network access or authentication. If the requested skill is not found, returns an error with available matching suggestions without modifying disk state.

When to use: Call before writing domain-specific code (game design, mobile QA, shaders, advanced refactoring) to load specialized instructions directly into context. When NOT to use: Do not use to search or discover unknown skills; use list_vault_skills first. Returns: Success confirmation, destination path, and the full markdown instructions of the equipped skill for immediate context execution.

Retrieves design specifications, color tokens, typography scales, elevation shadows, and layout rules for 74 digital brands from the local design library. Safe, read-only operation: performs local disk read with no side effects, mutations, network requests, or authentication requirements. If an unknown brand slug is requested, falls back to substring matching or returns an alphabetical index of available brands.

When to use: Call whenever styling UI components, design tokens, or pages to follow authentic brand aesthetics. When NOT to use: Do not use to calculate color accessibility compliance; use check_contrast to validate contrast ratios of retrieved colors. Returns: Full markdown design specification containing color palettes, typography, spacing, and component patterns; or a comma-separated index of all 74 brand names when brand='list'.

Computes mathematical WCAG 2.x relative luminance contrast ratio between foreground and background colors. Non-destructive, read-only computation with no side effects, network access, or authentication. Automatically normalizes 3-digit hex (#fff -> #ffffff), 6-digit hex, and 8-digit hex (ignoring alpha). On malformed hex input, returns a descriptive error specifying expected hexadecimal syntax without throwing unhandled exceptions.

When to use: Call whenever selecting color pairs for text, UI containers, buttons, or badges to verify WCAG AA compliance (>= 4.5:1 for normal body text, >= 3.0:1 for large or bold text >= 18pt). When NOT to use: Do not use for retrieving comprehensive brand styling or palettes; use get_brand_design instead. Returns: Mathematical ratio (e.g. 21.00:1), RGB conversions, PASS/FAIL verdicts for normal and large text, and luminance adjustment suggestions.

Inspects the operational presence and artifact health of the Graphify AST knowledge graph for a repository. Safe, read-only diagnostic: reads local metadata files with zero mutations, network calls, or side effects. If no graph exists (graphify-out/ missing), safely returns a non-error status stating the graph is unbuilt and provides the command to generate it.

When to use: Call before analyzing large codebases to determine if precomputed AST dependency graphs exist. When NOT to use: Do not use to query specific symbols, god classes, or module dependencies; use query_codebase_graph instead for graph traversal. Returns: Graph operational state (ACTIVE or NOT BUILT), artifact count, node/edge metrics if indexed, and actionable initialization instructions.

Traverses and searches the codebase AST knowledge graph to inspect symbol definitions, identify architectural god classes, trace dependencies, or examine functional clusters. Safe, read-only query: performs local file reads without mutations, network calls, or authentication. If the AST graph has not been precomputed (graphify-out/ missing), safely executes a non-failing structural fallback scan of top directories and code files with instructions to run graphify.

When to use: Call during codebase onboarding, architectural review, or refactoring to retrieve targeted structural relationships without loading massive source files. When NOT to use: Do not use if you only need high-level graph health or existence; call get_graphify_status instead. Returns: Structured architectural intelligence matching the query mode: summary statistics, ranked god classes by centrality, top connected dependencies, or matched symbol occurrences.