lasp/adamant-skills / adamant-tools
adamant-tools
Verifiedadamant-tools
Development tools for Adamant component workflows. Three standalone Python scripts that require no Docker or build environment (except adamant_inspect which reads generated files).
Tools
adamant_inspect.py -- Generated API Extractor
Purpose: After redo generates a component, extract a concise summary of the generated API: connectors, methods to override, helpers to call, events, data products, commands, parameters, faults, data dependencies, and tester histories.
When to use: After first successful redo build of a component, before writing the implementation body or test body. Eliminates guessing at generated method names, parameter types, and tester API.
python3 adamant_inspect.py <component_dir>
# Example:
python3 adamant_inspect.py src/components/voltage_monitor
Output sections:
- INIT PARAMETERS --
Initprocedure signature - SET_ID_BASES -- which ID bases the component requires
- RECV CONNECTORS -- abstract procedures to override (handler implementations)
- SEND/GET/REQUEST CONNECTORS -- helper procedures/functions to call
- DROPPED HANDLERS -- abstract null procedures for each send connector
- DATA DEPENDENCY HELPERS -- private
Get_*functions for data deps - DATA DEPENDENCY ABSTRACTS --
Get_Data_DependencyandInvalid_Data_Dependency - INSTANCE FIELDS --
Self.*record fields (Events, Data_Products, Faults, etc.) - EVENTS/DATA_PRODUCTS/COMMANDS/FAULTS -- creation function signatures
- PARAMETERS -- table fields and
Validate_Parameterssignature - TESTER API -- typed history names, connector histories, override fields
Key value: The #1 cold-start error is guessing at generated API names. This tool provides the exact signatures.
adamant_scaffold.py -- Component Scaffolder
Purpose: Generate all YAML models, implementation stubs (.ads/.adb), and test infrastructure (tests.yaml, env.py) from a spec file.
When to use: Starting a new component. Write a simple YAML spec describing what the component needs, then scaffold generates all files.
python3 adamant_scaffold.py <spec_file.yaml> [output_dir]
# Example:
python3 adamant_scaffold.py watchdog_timer_spec.yaml
# Creates src/components/watchdog_timer/ with all files
Spec file format:
name: watchdog_timer
description: Monitors heartbeats and triggers timeout faults
execution: passive
init:
- name: Timeout_Ms
type: Unsigned_32
description: Timeout period in milliseconds
connectors:
recv_sync:
- type: Tick.T
description: Periodic check
send:
- type: Event.T
description: Event output
- type: Data_Product.T
description: Data product output
get:
- type: Sys_Time.T
description: System time
events:
- name: Heartbeat_Received
description: Valid heartbeat detected
- name: Timeout_Triggered
description: Heartbeat timeout detected
param_type: Packed_U32.T
data_products:
- name: Heartbeat_Count
type: Packed_U32.T
description: Total heartbeats received
commands:
- name: Reset
description: Reset the watchdog timer
faults:
- name: Heartbeat_Timeout
description: No heartbeat within timeout period
parameters:
description: Configurable parameters
items:
- name: Timeout_Threshold
type: Packed_U32.T
default: "5000"
description: Timeout in milliseconds
data_dependencies:
description: External data dependencies
items:
- name: System_Health
type: Packed_Boolean.T
description: Overall system health status
Generated files:
<name>.component.yaml-- main component model<name>.events.yaml-- event definitions (if events specified)<name>.data_products.yaml-- data product definitions (if DPs specified)<name>.commands.yaml-- command definitions (if commands specified)<name>.parameters.yaml-- parameter definitions (if parameters specified)<name>.faults.yaml-- fault definitions (if faults specified)<name>.data_dependencies.yaml-- data dependency definitions (if data deps specified)component-<name>-implementation.ads-- implementation spec stubcomponent-<name>-implementation.adb-- implementation body stub with TODO markerstest/<name>.tests.yaml-- test configurationtest/env.py-- test environment
After scaffolding:
- Review and edit generated files (fill in TODOs)
admt build <component_dir>/build/src/component-<name>.ads(generate base class)- Run
adamant_inspect.pyto see the full generated API admt build <component_dir>/test/build/template/(generate test templates) -- or useadmt templates <component_dir>which runs this plus the stub copycp <component_dir>/test/build/template/*.ads <component_dir>/test/(skipped whenadmt templateshandles the copy)- Write test body, then
cd test && admt test
adamant_validate_yaml.py -- YAML Pre-flight Checker
Purpose: Fast structural validation of component YAML files without Docker. Catches the errors that most commonly cause build failures.
When to use: After writing or editing YAML files, before running admt build. Saves a Docker round-trip per error.
python3 adamant_validate_yaml.py <component_dir>
# Example:
python3 adamant_validate_yaml.py src/components/voltage_monitor
What it checks:
- Required fields in component YAML (description, execution, connector types/kinds)
- Framework component name collisions (58 known names)
- Init parameter structure and string defaults
- Feature YAML field names (
param_typenottypefor events,arg_typenottypefor commands) - Cross-file consistency (feature YAML exists but matching connector missing)
- File naming conventions (implementation spec/body, test directory structure)
- Test directory rules (env.py content, no .all_path in test dirs)
What it does NOT check:
- Full JSON Schema validation (requires Adamant Python environment)
- Ada compilation correctness
- Connector count or wiring validity
Integration with Skills
These tools complement the existing skills:
- adamant-component-dev: Use
adamant_scaffoldto create files,adamant_validate_yamlbefore building - adamant-testing: Use
adamant_inspectafter first build to see tester API before writing tests - adamant-build-system: Tools run outside Docker;
admt(and the underlyingredo) handle all in-container code generation
Environment Caching
admt owns a cached environment snapshot per registered project (/tmp/admt/<project>/env_snapshot.sh + exec.sh) and restores it in milliseconds on every call. The legacy adamant_env.sh wrapper still provides its own /tmp/.<project>_env_snapshot as a fallback.
# Run any in-container command through admt (snapshot-cached):
admt env exec "python3 /tmp/adamant_validate_yaml.py src/components/foo"
# Force snapshot regeneration after container recreation:
admt env refresh
Sub-agent workflow pattern:
# Step 1: Copy tools into container (once per container start)
docker cp adamant-tools/*.py <project>_container:/tmp/
# Step 2: Validate YAML (uses pykwalify if SCHEMAPATH set)
admt env exec "python3 /tmp/adamant_validate_yaml.py $COMP_DIR"
# Step 3: Build component
admt build $COMP_DIR/build/src/component-$NAME.ads
# Step 4: Inspect generated API
admt env exec "python3 /tmp/adamant_inspect.py $COMP_DIR"
# Step 5: Build and test
admt test $COMP_DIR/test
Fallback form (when admt is not applicable):
bash <project_dir>/docker/adamant_env.sh exec "python3 /tmp/adamant_validate_yaml.py src/components/foo"
pykwalify Schema Validation
Inside Docker (with source activate), the validator automatically uses Adamant's real pykwalify schemas from $SCHEMAPATH. This catches all YAML structural errors authoritatively -- no heuristic guessing.
Outside Docker, the validator falls back to heuristic checks (field names, connector consistency, naming conventions). These catch the most common errors but are not exhaustive.
Location
Tools are in the adamant-tools/ directory of the skills repository. Copy into Docker containers or run from the host against mounted volumes.
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| slack-gif-creatoranthropics/skillsKnowledge and utilities for creating animated GIFs optimized for Slack. Provides constraints, validation tools, and animation concepts. Use when users request animated GIFs for Slack like "make me a GIF of X doing Y for Slack." | Stars: 198k | Updated | Claude CodeCodex | Verified | Install |
| ponytailDietrichGebert/ponytailLazy senior dev mode for any coding task: the smallest change that fully solves it, and a reply a busy human understands in one read. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-debtDietrichGebert/ponytailHarvest every shortcut: comment into one debt ledger, so deferrals get tracked instead of forgotten. One-shot report. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-debtDietrichGebert/ponytailList every `shortcut:` comment (and older `ponytail:` ones) as a debt ledger. One-shot report, changes nothing. Use for "ponytail debt", "what did ponytail defer", "list the shortcuts", /ponytail-debt. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |