# Datalad Configuration

> >

- **Type:** Skill
- **Install:** `agentstack add skill-neuroanalytics-data-science-harness-datalad-configuration`
- **Verified:** Yes — security-reviewed for prompt injection and unsafe behavior
- **Seller:** [neuroanalytics](https://agentstack.voostack.com/s/neuroanalytics)
- **Installs:** 0
- **Category:** [Agent Skills](https://agentstack.voostack.com/c/agent-skills)
- **Latest version:** 0.1.0
- **License:** MIT
- **Upstream author:** [neuroanalytics](https://github.com/neuroanalytics)
- **Source:** https://github.com/neuroanalytics/data-science-harness/tree/main/plugins/datalad-cli/skills/datalad-configuration

## Install

```sh
agentstack add skill-neuroanalytics-data-science-harness-datalad-configuration
```

Requires the [AgentStack CLI](https://agentstack.voostack.com/docs/cli). Works with Claude Code, Cursor, and any MCP-compatible agent.

## About

# Skill: datalad-configuration

Read and write dataset-scoped, clone-local, or global DataLad/git configuration using
`datalad configuration`. This is the DataLad-aware wrapper around `git config` that
handles DataLad-specific keys and scopes correctly.

## Scopes

| Scope | Where stored | Effect |
|-------|-------------|--------|
| `dataset` | `.datalad/config` (committed) | Applies to all clones of this dataset |
| `local` | `.git/config` (not committed) | Applies only to this clone |
| `global` | `~/.gitconfig` | Applies to all repos for this user |

Use `dataset` scope for settings that should travel with the dataset (e.g., annex
backend choice). Use `local` for clone-specific overrides (e.g., credential paths).

## Steps

1. **Verify DataLad context** — check for `.datalad/` in the current directory:
   ```bash
   ls .datalad/ 2>/dev/null
   ```
   Configuration can still be inspected/set globally even outside a dataset, but warn
   the user if no dataset is found.

2. **Determine action** — from `$ARGUMENTS` or conversation context:
   - **get**: read the current value of a config key
   - **set**: write a new value
   - **unset**: remove a key

3. **Execute**:

   ### get
   ```bash
   datalad configuration get 
   # or list all keys in a section:
   datalad configuration get --scope dataset datalad.
   ```

   For common inspection use cases, also accept bare `git config`:
   ```bash
   git config --list --show-origin
   ```

   ### set
   Gather: key (in `section.key` form), value, and scope. Construct and show:
   ```bash
   datalad configuration set --scope  =
   ```

   Multiple keys can be set in one call:
   ```bash
   datalad configuration set \
     --scope dataset \
     "annex.backend=SHA256E" \
     "datalad.ui.progressbar=none"
   ```

   For `dataset` scope: the change writes to `.datalad/config`, which must be committed.
   Suggest: `datalad save -m "configure "` after setting.

   ### unset
   ```bash
   datalad configuration unset --scope  
   ```

4. **Show command and execute** — always display the full command before running.

## Common configuration keys

| Key | Scope | Purpose |
|-----|-------|---------|
| `annex.backend` | dataset | Hash backend for new files (`SHA256E` recommended) |
| `annex.largefiles` | dataset | Pattern deciding what goes into annex vs. git |
| `datalad.ui.progressbar` | local/global | Progress output style |
| `datalad.locations.cache` | local/global | Where DataLad caches data |
| `remote..annex-ignore` | local | Tell annex to ignore a remote |
| `user.name` / `user.email` | local/global | Git author identity |

## Constraints

- Always show the full command before executing.
- For `--scope dataset` changes: always remind the user to `datalad save` after setting —
  `.datalad/config` is a tracked file and the change is not recorded until committed.
- Never use `git config` directly for DataLad-specific keys (`datalad.*`, `annex.*`) —
  use `datalad configuration` to ensure DataLad's validation and defaults are applied.
- If the user asks to configure annex `largefiles` patterns, explain the gitattributes
  alternative (`echo "*.csv annex.largefiles=nothing" >> .gitattributes`) as it is more
  portable and does not require a dataset commit to take effect.

## Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

- **Author:** [neuroanalytics](https://github.com/neuroanalytics)
- **Source:** [neuroanalytics/data-science-harness](https://github.com/neuroanalytics/data-science-harness)
- **License:** MIT

Install and usage instructions live in the source repository linked above.

## Pricing

- **Free** — Free

## Security capabilities

Automated source analysis of v0.1.0 — what this tool can access:

- **Network access:** no
- **Filesystem access:** no
- **Shell / process execution:** no
- **Environment & secrets:** no
- **Dynamic code execution:** no

*"Yes" means the capability is present in the source — more access means more to trust, not that it is unsafe.*


## Versions

- **0.1.0** — security scan: passed — Imported from the upstream source.

## Links

- Listing page: https://agentstack.voostack.com/l/skill-neuroanalytics-data-science-harness-datalad-configuration
- Seller: https://agentstack.voostack.com/s/neuroanalytics
- Browse the marketplace: https://agentstack.voostack.com/browse

---
Listed on AgentStack — the marketplace for AI agent skills and MCP servers. Every listing is security-reviewed. Creators keep 70%.