Boosta Long to Shorts

Create, monitor, and troubleshoot Boosta API video-processing jobs from natural-language requests. Use this skill when a user asks to generate clips from a v...

Security Scan

The skill's code, instructions, and requested credential (BOOSTA_API_KEY) are consistent with its stated purpose of submitting and monitoring Boosta video-processing jobs; no disproportionate access, unexpected endpoints, or risky install behavior were found.

Purpose & Capability

Name/description match the actual behavior: the skill submits jobs to https://boosta.pro/api/v1, polls status, lists jobs, and checks usage. The only required environment variable is BOOSTA_API_KEY, which is appropriate for an API client. No unrelated credentials, binaries, or config paths are requested.

Instruction Scope

SKILL.md and bundled CLI both describe the same Boosta workflows and point at the provided script for deterministic execution. Minor inconsistency: SKILL.md's Decision Rules include logic to 'infer' video_type when the user doesn't provide one, but the bundled CLI enforces --video-type as required. This is a usability mismatch (not evidence of malice) and should be clarified by the author if inference is intended in agent-based invocations.

Install Mechanism

No install spec or remote downloads are included. The skill is instruction-plus-source: it bundles a small Python script (urllib-based) and reference docs. Nothing is fetched from arbitrary URLs or extracted to disk during install, so install risk is low.

Credentials

Only BOOSTA_API_KEY is required and used. The code reads that env var and uses it to set the Authorization header. No other secrets, system credentials, or config paths are requested. The script avoids printing the API key directly (SKILL.md explicitly warns against printing/storing keys).

Persistence & Privilege

The skill is not always-enabled and does not request persistent system-level privileges. It does not modify other skills or global agent configs. Autonomous invocation is allowed by default (platform normal), but there are no additional privilege escalations requested by the skill.

Assessment

This skill appears coherent and limited to interacting with Boosta's API. Before installing, confirm you trust boosta.pro and that your BOOSTA_API_KEY is scoped appropriately (use a key with limited permissions/credits if possible). Note the small mismatch: the SKILL.md mentions inferring video_type when absent, but the bundled CLI requires --video-type; clarify whether the agent should infer this before calling the CLI. Test the skill with a non-production/low-credit account first to verify behavior, and rotate/revoke the API key if you stop using the integration. Finally, be aware that the agent can invoke this skill autonomously (normal platform behavior); if you want to prevent autonomous runs, adjust agent/skill permissions in your environment.

Boosta Video Api

Overview

Use this skill to execute end-to-end Boosta API workflows: submit job, poll until completion, return clip URLs, and handle common API failures. Prefer the bundled script for deterministic execution and consistent error handling.

Quick Start Workflow

1. Validate required inputs:

• video_url (required)
• video_type (required, see references/video-types.md)
• config_name (optional)

2. Validate credentials:

• Require BOOSTA_API_KEY in environment.
• Never print or store raw API keys in output.

3. Submit job:

• Endpoint: POST /api/v1/jobs
• Base URL: https://boosta.pro/api/v1
• Body: video_url, video_type, optional config_name

4. Poll status:

• Endpoint: GET /api/v1/jobs/:job_id
• Stop when status=completed or status=failed.
• If completed, return clip_urls.

5. Handle API errors:

• 401: invalid or missing key
• 400: invalid payload (missing video_url, invalid video_type)
• 403: no credits
• 429: rate limited (respect retry_after)
• active_job_exists: reuse returned job_id and continue polling

Commands

Use the bundled script:

python3 scripts/boosta_job.py --help

Submit and wait for completion:

export BOOSTA_API_KEY="sk_live_..."
python3 scripts/boosta_job.py submit \
  --video-url "https://youtube.com/watch?v=xxx" \
  --video-type "conversation" \
  --config-name "My Config" \
  --wait

Check status:

python3 scripts/boosta_job.py status --job-id "job_1234567890_abc123"

List completed jobs:

python3 scripts/boosta_job.py list

Check usage:

python3 scripts/boosta_job.py usage

Output Contract

When user asks to create clips, return:

• job_id
• final status
• clips_count if present
• clip_urls when completed
• clear next step if processing/failed

When user asks only to check status, return:

• job_id
• current status
• progress/step if present

Decision Rules

• If user does not provide video_type, infer it with references/video-types.md and state inference explicitly.
• If API returns active_job_exists, continue with provided job_id instead of creating another job.
• On 429, wait retry_after seconds (fallback to 60 if missing) and retry.
• Keep polling interval between 10 and 20 seconds to avoid spam.
• Avoid parallel job submission per key because API allows one active job at a time.

References

• API overview and endpoint contracts: references/api-reference.md
• Video type selection: references/video-types.md
• Error handling playbook: references/errors.md