System Skills

Utility Tools

Name Description
file-read Read file. Path traversal prevention. Supports line ranges and pattern search with context.
file-write Write file with auto-mkdir. Path traversal prevention. Supports create, append, str_replace, and insert commands.
file-list List directory. Path traversal prevention. Supports recursive depth and glob filtering.
file-glob Find files by glob pattern. Returns { totalFiles, truncated, filePaths }. Supports maxDepth and limit. Skips common non-project directories.
file-search Regex search across project files. Returns matches with file paths, line numbers, and content. Supports include glob filter, contextLines, and three output modes: content, files, count.
web-fetch Fetch a URL and extract text content. Converts HTML to clean text. Three modes: selective, truncated, full. Returns { url, title, content, truncated }.
shell Execute a bash command. Returns { stdout, stderr, exitCode }. Configurable working directory and timeout.
code Evaluate JavaScript code in the current process with ctx in scope. Supports await.
invoke Dynamically invoke any tool by name.
skill-list List available skills. Supports query, tags filter, and includeHidden.
skill-describe Return full skill details: params, returns, tags, allowed-tools, source, authority, precedence.
skill-register Register a new skill at runtime from code or markdown source. Hidden from agents.
skill-unregister Remove a dynamically registered skill by ref. Hidden from agents.
cache-clear Clear internal caches. Pass ref to clear one tool, or omit for full wipe.

For events, callbacks, session, schedule, and state tools, see Runtime. For gateway, email, Slack, trust, and roles tools, see Messaging. For hub, deploy, and sandbox tools, see Operations.

Debug & Profile

Pipeline observability tools activated via metadata keys. Debug, profile, and observe are not part of the default pipeline — they use $hook annotations to bind as hooks on the pipeline when their metadata keys appear in a skill's cascade.

Activation

Add the metadata key to a skill's cascade to activate:

# Enable debug hooks via metadata
metadata:
  debug: true

# Enable profile hooks via metadata
metadata:
  profile: true

debug

Observe pipeline execution, set breakpoints, and step through chain entries. When activated via hooks, emits pipeline-enter, pipeline-exit, next-enter, next-exit, and breakpoint-hit events to the debug topic. Supports breakpoints, stepping, include/exclude filters, verbose snapshots, and override-on-continue. Can be disabled per-context via nonlocals.debug = false.

# Via CLI wrapper
agent-apps skill-debug --ref hello
agent-apps skill-debug --ref hello --breakpoints '["execute"]'
agent-apps skill-debug --ref hello --verbose true

profile

Collect timing data for every pipeline invocation and chain entry. Records { phase, name, tool, contextId, rootContextId, depth, start, duration } for each entry. Data collected in-memory, scoped by rootContextId, retrieved via profile { report: true }. Can be disabled per-context via nonlocals.profile = false.

# Via CLI wrapper
agent-apps skill-profile --ref hello

Help

Show GNU-style usage documentation for any skill. Formats schema properties as CLI flags with types, defaults, and required markers. Includes fuzzy matching for "did you mean?" suggestions when a skill is not found.

agent-apps help --ref shell
agent-apps help --ref file-read

Ask AI