MCP Facade Action Reference

AUTO-GENERATED — do not edit by hand. Regenerate with uv run python scripts/generate_facade_actions_doc.py. Drift-gated by tests/unit/docs/test_facade_actions_doc_drift.py (regenerates in-memory and diffs).

The MCP server exposes 8 facade tools routing 73 actions via the action parameter. This reference is generated from the live facade registry (tree_sitter_analyzer/mcp/_tool_registry.py) and each inner tool's inputSchema — the same schema the runtime strict-parameter guard enforces, so a wrong param guess in this table would fail at runtime too (and vice versa).

Reading the tables:

• Params — accepted top-level parameters; * marks required ones. Facades mechanically alias the canonical symbol onto inner function_name/class_name params (noted inline). Every facade also accepts action (required) itself.
• Response keys — the statically declared ToolResponse envelope (get_output_schema()); * marks guaranteed keys. error appears on failures. "+ action payload" means the action layers its own result keys on top (additionalProperties: true); payload shapes are not statically declared, so they are not listed here — see the facade description for per-action semantics.
• CLI twin — the CLI flag (or console script) covering the same capability, from the CLI-parity contract. 4 actions have no authoritative CLI mapping and show — (honest gap, not an omission).
• Bespoke routes (closures with hand-rolled arg handling, e.g. nav action=test_map) have their params pinned in the generator with source provenance; the generator fails if the live route set drifts from those pins.

search — 9 actions

Action	Params (required *)	Response keys (top-level)	CLI twin
batch	queries*	success, verdict, agent_summary, error + action payload	--batch-search
chain	query*, compact, include_code, max_files, max_symbols, output_format	success, verdict, agent_summary, error + action payload	--codegraph-query
content	query*, case, context, context_after, context_before, count_only_matches, enable_parallel, encoding, exclude_globs, exclude_types, extensions, file_types, files, files_with_matches, fixed_strings, follow_symlinks, group_by_file, hidden, include_globs, include_stats, invert_match, max_count, max_depth, max_filesize, multiline, no_ignore, only_matching, optimize_paths, output_file, output_format, pcre2, roots, sort, summary_only, suppress_output, timeout_ms, total_only, word	success, verdict, agent_summary, error + action payload	search-content (console script)
grep	query, roots, case, changed_before, changed_within, context_after, context_before, count_only_matches, depth, encoding, exclude, exclude_globs, extensions, file_limit, fixed_strings, follow_symlinks, full_path_match, glob, group_by_file, hidden, include_globs, max_count, max_filesize, multiline, no_ignore, optimize_paths, output_file, output_format, pattern, size, sort, summary_only, suppress_output, timeout_ms, total_only, types, word	success, verdict, agent_summary, error + action payload	find-and-grep (console script)
query	file_path, filter, find_references, language, max_count, output_file, output_format, query_key, query_string, result_format, suppress_output, symbol, symbol_type — requires file_path or symbol; file-scoped queries take exactly one of query_key/query_string	success, verdict, agent_summary, error + action payload	--query-key
select	selector*, max_results, output_format	success, verdict, agent_summary, error + action payload	—
subscribe	selector*, min_interval, output_format	success, verdict, agent_summary, error + action payload	—
symbol	query*, kind, language, limit, output_format	success, verdict, agent_summary, error + action payload	--symbol-search
unsubscribe	output_format, selector, sub_id	success, verdict, agent_summary, error + action payload	—

nav — 14 actions

Action	Params (required *)	Response keys (top-level)	CLI twin
call_path	source_function, target_function, direction, max_depth, max_paths, output_format, source_file, target_file	success, verdict, agent_summary, error + action payload	--call-path
callee_tree	symbol*, file_path, max_depth, max_nodes, output_format	success, verdict, agent_summary, error + action payload	--callee-tree
callees	function_name* (or symbol as alias), scope (point|graph, default point), file_path, limit (scope=point), depth (scope=graph), output_format	success, verdict, agent_summary, error + action payload	--call-graph
caller_tree	symbol*, file_path, max_depth, max_nodes, output_format	success, verdict, agent_summary, error + action payload	--caller-tree
callers	function_name* (or symbol as alias), scope (point|graph, default point), file_path, limit (scope=point), depth (scope=graph), output_format	success, verdict, agent_summary, error + action payload	--call-graph
co_change	symbol or file_path (one required), max_commits (default 500), min_shared (default 3), max_results (default 20), output_format	success, verdict, agent_summary, error + action payload	--co-change
context	task* (or symbol/query as alias), max_nodes, max_code_blocks, include_graph, output_format	success, verdict, agent_summary, error + action payload	--codegraph-context
impact	mode*, depth, file_path, function_name (symbol aliases function_name), function_names, include_tests, output_format	success, verdict, agent_summary, error + action payload	--codegraph-impact
lineage	symbol*, file_paths, max_depth, output_format	success, verdict, agent_summary, error + action payload	--symbol-lineage
navigate	symbol*, depth, file_path, mode, output_format	success, verdict, agent_summary, error + action payload	--codegraph-navigate
resolve	symbol*, mode, output_format	success, verdict, agent_summary, error + action payload	--symbol-resolve
test_map	symbol* (or function_name as alias), file_path, output_format	success, verdict, agent_summary, error + action payload	--test-map
trace	symbol*, case_sensitive, exclude_patterns, file_path, max_results, project_root, word_match	success, verdict, agent_summary, error + action payload	--trace-impact
xref	file_path, include_callees, include_callers, include_file_deps, include_imports, mode, output_format, symbol	success, verdict, agent_summary, error + action payload	--codegraph-xref

structure — 9 actions

Action	Params (required *)	Response keys (top-level)	CLI twin
analyze	file_path*, format_type, language, output_file, output_format, suppress_output	success, verdict, agent_summary, error + action payload	--structure
ast_path	file_path*, language, line, max_depth, mode, output_format	success, verdict, agent_summary, error + action payload	--ast-path
class_detail	class_name* (or query/symbol as alias), language, output_format	success, verdict, agent_summary, error + action payload	--class-inspect
class_tree	class_name (symbol aliases class_name), max_depth, mode, output_format	success, verdict, agent_summary, error + action payload	--class-hierarchy
explore	includeCode, maxFiles, maxSymbols, output_format, query, symbol, symbols	success, verdict, agent_summary, error + action payload	--codegraph-explore
outline	file_path*, include_fields, include_imports, language, listed_cap, output_format	success, verdict, agent_summary, error + action payload	--outline
read	single: file_path* + start_line* [+ end_line, start_column, end_column, format, output_file, suppress_output, output_format, allow_truncate, fail_fast]; batch: requests*	success, verdict, agent_summary, error + action payload	--partial-read
signatures	file_path*, language, output_format	success, verdict, agent_summary, error + action payload	—
sitemap	directory, language, max_files, max_symbols, mode, output_format	success, verdict, agent_summary, error + action payload	--codegraph-sitemap

health — 12 actions

Action	Params (required *)	Response keys (top-level)	CLI twin
dead	include_test_files, max_dead, max_imports, max_variables, mode, output_format, path	success, verdict, agent_summary, error + action payload	--dead-code
deps	file_path, mode, output_format	success, verdict, agent_summary, error + action payload	--dependencies
file	file_path*, compact_only, language, output_format	success, verdict, agent_summary, error + action payload	--file-health
heatmap	directory, file_path, function_name (symbol aliases function_name), language, max_files, mode, output_format	success, verdict, agent_summary, error + action payload	--codegraph-complexity-heatmap
imports	file_path, max_depth, mode, output_format	success, verdict, agent_summary, error + action payload	--import-graph
matrix	mode*, file_path, output_format, threshold, top_k	success, verdict, agent_summary, error + action payload	--dependency-matrix
overview	max_coupled_files, max_dead, max_entry_points, max_hubs, output_format	success, verdict, agent_summary, error + action payload	--codegraph-overview
patterns	file_path*, categories, output_format, severity_threshold	success, verdict, agent_summary, error + action payload	--code-patterns
project	compact_only, max_files, min_grade, output_format	success, verdict, agent_summary, error + action payload	--project-health
routes	file_path, framework, mode, output_format, url_pattern	success, verdict, agent_summary, error + action payload	--detect-routes
scale	file_path, file_paths, include_complexity, include_details, include_guidance, language, metrics_only, output_format	success, verdict, agent_summary, error + action payload	--metrics-only
test_gap	coverage_json, file_path, include_covered, language, max_files, max_gaps, mode, output_format	success, verdict, agent_summary, error + action payload	--test-gap

edit — 8 actions

Action	Params (required *)	Response keys (top-level)	CLI twin
ast_diff	file_path, include_node_bodies, language, mode, new_file, new_ref, new_source, old_file, old_ref, old_source, output_format	success, verdict, agent_summary, error + action payload	--ast-diff
classify	file_path, hunk_cap, include_ast_nodes, language, mode, new_ref, new_source, old_ref, old_source, output_format	success, verdict, agent_summary, error + action payload	--semantic-classify
constraints	output_format, path_filter, severity_min	success, verdict, agent_summary, error + action payload	--check-constraints
guard	modification_type, symbol, file_path	success, verdict, agent_summary, error + action payload	--modification-guard
impact	agent_summary_only, compact_only, include_tests, mode, output_format, pr_url, resource_profile, scope_mode, scope_paths	success, verdict, agent_summary, error + action payload	--change-impact
pr	include_call_graph, mode, output_format, pr_url	success, verdict, agent_summary, error + action payload	--pr-review
refactor	file_path*, include_extractions, include_skeleton, language, max_suggestions, output_format	success, verdict, agent_summary, error + action payload	--refactor
safe	file_path*, compact_only, edit_type, output_format	success, verdict, agent_summary, error + action payload	--safe-to-edit

project — 10 actions

Action	Params (required *)	Response keys (top-level)	CLI twin
doc_sync	doc_patterns, output_format	success, verdict, agent_summary, error + action payload	--doc-sync
files	absolute, changed_before, changed_within, count_only, depth, exclude, extensions, follow_symlinks, full_path_match, glob, hidden, limit, min_depth, no_ignore, one_file_system, output_file, output_format, path, pattern, prune, roots, show_errors, size, strip_cwd_prefix, suppress_output, threads, types	success, verdict, agent_summary, error + action payload	list-files (console script)
journal	alternatives, id, limit, mode, new_id, output_format, path_scope, query, rationale, related_symbols, scope_paths, tags, title, verdict, verdict_filter	success, verdict, agent_summary, error + action payload	--decision-journal
metrics	output_format, sections	success, verdict, agent_summary, error + action payload	--codegraph-metrics
overview	include_health, max_depth, output_format	success, verdict, agent_summary, error + action payload	--overview
parser	include_supported, language, output_format	success, verdict, agent_summary, error + action payload	--parser-readiness
skills	output_format, skills_root	success, verdict, agent_summary, error + action payload	--agent-skills
smart	file_path*, output_format	success, verdict, agent_summary, error + action payload	--smart-context
tools	(none)	success, verdict, agent_summary, error + action payload	--check-tools
workflow	output_format, target_path	success, verdict, agent_summary, error + action payload	--agent-workflow

index — 7 actions

Action	Params (required *)	Response keys (top-level)	CLI twin
auto	max_files, mode, output_format	success, verdict, agent_summary, error + action payload	--autoindex
build	add_notes, roots	success, verdict, agent_summary, error + action payload	--build-project-index
cache	backend, file_path, force, include_activation, language, limit, max_files, mode, poll_interval, query, symbol	success, verdict, agent_summary, error + action payload	--ast-cache
full	include_activation, max_files, mode, output_format, resolve_synapse	success, verdict, agent_summary, error + action payload	--full-index
knowledge	backend, include_docs, max_edges, max_files, max_nodes, mode, output_format	success, verdict, agent_summary, error + action payload	--knowledge-graph-index
status	include_lag, output_format	success, verdict, agent_summary, error + action payload	--codegraph-status
sync	max_files, mode, output_format	success, verdict, agent_summary, error + action payload	--incremental-sync

viz — 4 actions

Action	Params (required *)	Response keys (top-level)	CLI twin
graph	depth, direction, file_path, function, max_edges, mode, output_format, visualization_format	success, verdict, agent_summary, error + action payload	--codegraph-visualize
knowledge	export_format, focus, lod, max_edges, max_nodes, output_format, uml_kind	success, verdict, agent_summary, error + action payload	--knowledge-graph-export
similarity	include_bodies, max_groups, min_group_size, min_lines, mode, output_format, path_filter, use_cache	success, verdict, agent_summary, error + action payload	--code-similarity
uml	class_name (symbol aliases class_name), diagram, file_path, function_name (symbol aliases function_name), include_external_bases, include_tests, max_depth, max_edges, max_nodes, max_paths, output_format, package_depth, source, target	success, verdict, agent_summary, error + action payload	--uml