Aiwg run

Execute a named script defined in .aiwg/aiwg.config or list all available scripts

install

source · Clone the upstream repo

git clone https://github.com/jmagly/aiwg

Claude Code · Install into ~/.claude/skills/

T=$(mktemp -d) && git clone --depth=1 https://github.com/jmagly/aiwg "$T" && mkdir -p ~/.claude/skills && cp -r "$T/.agents/skills/run" ~/.claude/skills/jmagly-aiwg-run && rm -rf "$T"

manifest: .agents/skills/run/SKILL.md

source content

Run

You execute a named script defined in

.aiwg/aiwg.config

, or list all available scripts when no name is given.

Triggers

Alternate expressions and non-obvious activations (primary phrases are matched automatically from the skill description):

"what scripts are available" → list scripts
"run the deploy script" → run
```
deploy
```
"execute test script" → run
```
test
```
"aiwg run" → list if no name; run if name given

Trigger Patterns Reference

Pattern	Example	Action
Named run	"run the test script"	Run `aiwg run test`
Explicit list	"list aiwg scripts"	Run `aiwg run` (no args)
With args	"run deploy with flag --prod"	Run `aiwg run deploy --prod`
Unknown name	"run the build script"	Run `aiwg run build` ; report if not found

Behavior

When triggered:

Verify config exists:
- Read
```
.aiwg/aiwg.config
```
  to confirm
```
scripts
```
  block is present.
- If no config found, suggest
```
aiwg init
```
  to create one.
Extract arguments:
- Is a script name specified? If yes, run it.
- If no name, list all registered scripts with their shell commands.
- Are additional arguments passed through? Append them to the shell invocation.

Run the appropriate command:

# List all available scripts
aiwg run

# Run a named script
aiwg run <script-name>

# Run with pass-through arguments
aiwg run <script-name> -- <extra-args>

What "run" does:
- Reads the script's shell command from
```
.aiwg/aiwg.config
```
- Executes it in the project root
- Streams output to the terminal
- Reports exit code on completion
Report the result — show the command that was executed, stream output, and confirm success or failure with exit code.

Examples

Example 1: List available scripts

User: "What scripts do I have registered in aiwg?"

Extraction: List request, no script name

Action:

aiwg run

Response:

Registered scripts (from .aiwg/aiwg.config):

  test    → npm test
  lint    → npm run lint
  deploy  → ./scripts/deploy.sh --env staging

Run a script with: aiwg run <script-name>

Example 2: Run a named script

User: "Run the lint script"

Extraction: Script name

lint

Action:

aiwg run lint

Response: Executes

npm run lint

, streams output, then reports: "Script

lint

completed (exit 0)."

Example 3: Run with pass-through arguments

User: "Run the deploy script with --env production"

Extraction: Script name

deploy

, extra args

--env production

Action:

aiwg run deploy -- --env production

Response: Executes

./scripts/deploy.sh --env staging --env production

(or however the script merges args), then reports exit status.

Example 4: Script not found

User: "Run the build script"

Extraction: Script name

build

, not present in config

Action:

aiwg run build

Response: "No script named

build

found in

.aiwg/aiwg.config

. Available scripts: test, lint, deploy. Register

build

with

aiwg init --interactive

or edit

.aiwg/aiwg.config

directly."

Example 5: No config present

User: "aiwg run test"

Extraction: No

.aiwg/aiwg.config

found

Response: "

.aiwg/aiwg.config

not found. Run

aiwg init

to create one and register scripts."

Clarification Prompts

If the user's intent is ambiguous:

"Which script would you like to run? Available: test, lint, deploy."
"No script name given — should I list all registered scripts?"

References

@$AIWG_ROOT/src/cli/handlers/run.ts —
```
run
```
command handler
@$AIWG_ROOT/docs/cli-reference.md — CLI reference
@$AIWG_ROOT/agentic/code/addons/aiwg-utils/skills/init/SKILL.md — Registering scripts via init