# Orchestrate Multi-Agent Product Delivery

Complex product work can require coordination across planning, coding, testing, and operations. Loki Mode provides a structured multi-agent workflow with task queues, checkpoints, and review steps.

## Install

```bash
npx skillstore add sickn33/loki-mode
```

## Metadata

- Status: approved
- Slug: sickn33-loki-mode
- Skillstore revision: r1
- Version status: missing
- Tree hash: bbf0b6d1c25f3f41c3d5ebd3cf02dcf2e8ce3ea62c4555356121212087a8c24f
- Author: sickn33
- GitHub username: sickn33
- License: MIT
- Repository: https://github.com/sickn33/antigravity-awesome-skills/tree/main/skills/loki-mode
- Ref: c7830d47eaa856f41e62493881df761fc6900356
- Supported tools: Claude, Codex, Claude Code
- Audit status: complete
- Agent install advisory: blocked
- Manual install advisory: allowed\_with\_warning
- Artifact signature: available
- Audit attestation: unavailable
- Human verification: not\_verified
- Risk factors: external\_commands, network, filesystem, scripts, env\_access
- Quality score: 38
- Quality tier: warning
- Public page: https://skillstore.pages.dev/skills/sickn33-loki-mode
- Manifest: https://skillstore.pages.dev/api/skills/sickn33-loki-mode/manifest

## Capabilities

- Organizes work into product delivery phases from discovery through deployment.
- Tracks tasks and workflow state in a local .loki workspace.
- Provides RARV reasoning, action, reflection, and verification guidance.
- Describes parallel agent dispatch and specialist review patterns.
- Includes resource monitoring, audit logging, and configurable command blocking.
- Generates a local dashboard and screenshots for workflow demonstrations.

## Use Cases

- Plan a new service: Turn a product brief into phased architecture, development, testing, and release tasks.
- Coordinate release readiness: Organize review, test, documentation, and deployment preparation tasks in one workflow.
- Evaluate agent workflows: Use the local dashboard and benchmark materials to study multi-agent coordination patterns.

## Prompt Templates

### Start a small workflow

```
Loki Mode, create a plan for a small API feature. Stop after the plan and ask for approval before making changes.
```

### Prepare release tasks

```
Loki Mode, organize release-readiness tasks for this repository. Include tests, documentation, and a human approval checkpoint.
```

### Coordinate a feature

```
Loki Mode, break this approved feature into scoped implementation and review tasks. Keep all changes inside the repository and require approval before deployment.
```

### Review workflow controls

```
Loki Mode, audit this workflow plan for unsafe permissions, external actions, and missing approval gates. Produce recommendations only.
```

## Limitations

- It requires significant human review before any deployment or external action.
- It does not guarantee a deployed product, revenue, or zero-intervention execution.
- It may consume substantial local compute when many agents run in parallel.
- The included screenshot script weakens Chromium sandbox protections.

## Best Practices

- Keep workflows scoped to a repository and approved objective.
- Require human approval before deployments, external messages, or permission changes.
- Set conservative agent limits and verify each completed task.

## Anti Patterns

- Do not run with unrestricted permissions for routine work.
- Do not allow agents to deploy or contact services without approval.
- Do not treat generated plans or benchmarks as proof of production readiness.

## Security Audit

- Audited at: 2026-07-21T08:15:23.985\+00:00
- Summary: Static destructive-command, screen-capture, weak-cryptography, environment, and long-line alerts are false positives: they are blocklist text, documentation, local demo tooling, or benchmark artifacts. However, the skill explicitly requires a permission bypass and directs unsupervised execution without confirmations; it also contains a dashboard HTML injection risk. Publishing should remain blocked until these issues are remediated. Static review was capped at 400/5857 representative findings; omitted static matches are unconfirmed, so automatic publishing stays disabled until manual review.

## Stats

- Views: 109
- Downloads: 21
- Favorites: 0
- Popularity score: 0