Pain Point Finder

Discover pain points, frustrations, and unmet needs on Reddit using PullPush API. No API keys required. Use to find startup ideas backed by real user complai...

Security Scan

The skill's code, documentation, and runtime requirements are consistent with its stated purpose (finding Reddit pain points via the PullPush API); it requires only node and makes unauthenticated requests to api.pullpush.io.

Purpose & Capability

Name/description (Reddit pain-point discovery via PullPush) match the code and docs: the script queries api.pullpush.io for submissions and comments and computes pain/validation metrics. Required binary (node) is appropriate and no unrelated credentials or services are requested.

Instruction Scope

SKILL.md instructs running the included node script with discover/scan/deep-dive subcommands. The runtime instructions and script only reference PullPush API calls, local JSON output, and stderr logging. There is no instruction to read unrelated local files or to exfiltrate secrets.

Install Mechanism

No install spec (instruction-only) and the bundled code is self-contained. There are no downloads from untrusted URLs or archive extraction steps. Requiring node is proportional.

Credentials

No environment variables or credentials are required. The script makes unauthenticated requests to a single third-party host (api.pullpush.io), which is appropriate for the stated design.

Persistence & Privilege

The skill is not set to always: true and does not request system-wide changes or other skills' credentials. It runs on demand and only performs network queries to PullPush.

Assessment

This skill appears internally consistent with its purpose. Before installing, consider: (1) it will make many network calls to api.pullpush.io (self-limited to ~300/run) so confirm you’re comfortable with that third-party service and its terms/privacy; (2) outputs include Reddit quotes which can contain personal information — review/sanitize before sharing; (3) the code is bundled and runs locally under node (no hidden installers), but if you want extra assurance, inspect the full script (the truncated portion shown is consistent with behavior described) to confirm there are no other external endpoints or file I/O. If you rely on an enterprise environment with data controls, verify that outbound requests to api.pullpush.io are acceptable.

Discover validated pain points on Reddit. Searches for frustrations, complaints, and unmet needs, then analyzes comment threads for agreement signals and failed solutions. Powered by PullPush API — no API keys needed.

Workflow

Follow these 4 phases in order. Each phase builds on the previous.

Phase 1: Discover Subreddits

Find the right subreddits for the user's domain.

node {baseDir}/scripts/pain-points.mjs discover --domain "<user's domain>" --limit 8

Example:

node {baseDir}/scripts/pain-points.mjs discover --domain "project management" --limit 8

Take the top 3-5 subreddits from the output for phase 2.

Phase 2: Scan for Pain Points

Broad search across discovered subreddits.

node {baseDir}/scripts/pain-points.mjs scan \
  --subreddits "<sub1>,<sub2>,<sub3>" \
  --domain "<domain>" \
  --days 90 \
  --limit 20

Example:

node {baseDir}/scripts/pain-points.mjs scan \
  --subreddits "projectmanagement,SaaS,smallbusiness" \
  --domain "project management" \
  --days 90 \
  --limit 20

Review the scored posts. Posts with high painScore and high num_comments are the best candidates for deep analysis.

Phase 3: Deep-Dive Analysis

Analyze comment threads of top posts for agreement and solution signals.

Single post:

node {baseDir}/scripts/pain-points.mjs deep-dive --post <post_id>

Top N from scan output:

node {baseDir}/scripts/pain-points.mjs deep-dive --from-scan <scan_output.json> --top 5

Look at the validationStrength field:

• strong: widespread, validated pain (agreementRatio > 0.20, 10+ agreements)
• moderate: notable pain with some validation
• weak: some signal but limited agreement
• anecdotal: one person's complaint, needs more evidence

Phase 4: Synthesis (you do this)

For each validated pain point, present a structured proposal:

1. Problem: One-sentence description of the pain
2. Evidence: Top quotes + agreement count + subreddit
3. Who feels this: Type of person/business affected
4. Current solutions & gaps: What people have tried (from solutionAttempts) and why it fails
5. Competitive landscape: Tools mentioned (from mentionedTools)
6. Opportunity: What's missing in current solutions
7. Idea sketch: Brief product/service concept
8. Validation: strong/moderate/weak + data backing it

Options Reference

discover

Flag	Default	Description
--domain	required	Domain to explore
--limit	10	Max subreddits to return

scan

Flag	Default	Description
--subreddits	required	Comma-separated subreddit list
--domain		Domain for extra search queries
--days	365	How far back to search
--minScore	1	Min post score filter
--minComments	3	Min comment count filter
--limit	30	Max posts to return
--pages	2	Pages per query (more = deeper, slower)

deep-dive

Flag	Default	Description
--post		Single post ID or Reddit URL
--from-scan		Path to scan output JSON
--stdin		Read scan JSON from stdin
--top	10	How many posts to analyze from scan
--maxComments	200	Max comments to fetch per post

Rate Limits

The script self-limits to 1 request/sec, 30/min, 300/run. If PullPush is slow or returns errors, it retries with exponential backoff. Progress is logged to stderr.

Tips

• Start broad with --days 90 then narrow to --days 30 for recent trends
• High num_comments + high score = validated pain (many people agree)
• High painScore + low num_comments = niche pain (worth investigating)
• The mentionedTools in deep-dive output maps the competitive landscape
• Posts with validationStrength: "strong" are the best startup candidates