SaaS Idea Scout
ReviewAudited by ClawScan on May 10, 2026.
Overview
The skill is a coherent SaaS research workflow, but it uses an autonomous cron watchdog and retrying sub-agent swarm that could keep relaunching work without a visible retry limit.
Use this only if you are comfortable with a 30–45 minute multi-agent research run. Before starting, confirm how to cancel the run, clean up the cron watchdog, and limit retries; avoid entering confidential founder or product information unless it is acceptable for web research and sub-agent processing.
Findings (4)
Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.
If a run stalls or is interrupted, background recovery behavior could continue waking or managing the workflow until it is cleaned up.
The skill intentionally creates a background watchdog that can wake the orchestrator during a multi-phase run. This is disclosed and purpose-related, but it is a persistence mechanism that needs clear cleanup and stop behavior.
A cron watchdog spans all spawn phases... It is created after Phase 1 and removed after Phase 4 completes.
Before using it, confirm how to stop the watchdog and remove any cron entry if the run fails or is canceled.
A bad output, missing file, or verification mismatch could cause repeated agent runs, consuming time, tokens, or other resources.
The retry loop is described as continuing until files pass verification, but the visible artifact does not show a retry cap, timeout, budget limit, or user approval checkpoint for repeated re-spawns.
If files are missing: identify which agents, re-spawn individually, yield again... Repeat until all files for the current batch exist
Add or verify a maximum retry count, timeout, and user confirmation before additional recovery attempts.
The agent will run local verification commands during the pipeline.
The skill uses local shell commands such as stat and grep to verify generated markdown files. The shown commands are scoped to output checks and are not destructive, but users should know local command execution is part of the workflow.
On every wake, verify expected output files exist using exec commands
Run it in a workspace where local file checks are acceptable, and avoid giving it paths outside the intended run directory.
Confidential startup ideas, founder details, or market strategy could be included in sub-agent prompts or external research queries.
User-provided business and founder context is passed into sub-agent tasks and used with web_search/web_fetch for research. This is expected for the skill's purpose, but it means sensitive business details may be shared across agent contexts and search queries.
Domain context (industry, constraints, target customers, founder profile — from the user)
Avoid entering confidential or proprietary details unless you are comfortable with them being used for multi-agent and web-research workflows.