More screenshots

    Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIDeepSeekDeepSeekMuseMuse+22 more

    🩹 Customer Pain Extractor

    3

    Flag and group keyword-based candidate pain signals in supplied feedback, with quotes and text-occurrence counts for human review. Select CSV feedback columns explicitly. Counts are not respondents or proof of demand.

    $12

    Secure checkout via Stripe

    Money-back guarantee

    Converts to your local currency at checkout

    Security scanned
    🩹 Customer Pain Extractor

    🩹 Customer Pain Extractor

    Example session with this skill installed

    Run the bundled synthetic fixtures/risky-sample/feedback.txt and show the cue counts and quotes. Do not interpret the counts as people or demand.

    • Read your context and instructions
    • Compiled the customer pain extractor

    Actual synthetic fixture run

    This example uses invented feedback bundled with the package, not customer evidence.

    
    # Customer pain extractor
    
    Found 5 matched statements with 7 rule matches.
    
    Counts describe supplied text occurrences, not unique customers, buyers or proven demand.
    
    ## 1. switching risk (2 mentions, score 3)
    - "I cancelled because the import was broken every time."
    - "We found a better tool after the app stayed too slow."
    
    ## 2. ease of use (1 mentions, score 1)
    - "The setup is confusing and I can't tell what to do next."
    
    ## 3. missing feature (1 mentions, score 1)
    - "I wish it had a simple weekly report for my team."
    
    ## 4. price and value (1 mentions, score 1)
    - "The pricing feels too expensive for a tiny project."
    

    Review the original context before accepting any cue as a complaint. These counts do not establish five respondents, willingness to pay or a market opportunity. Scores are within-sentence keyword weights, not a 1–10 severity rating.

    Connects securely to your tools. The creator never sees your data.

    About this skill

    Review feedback with traceable cue counts

    Customer Pain Extractor reads authorized local text, Markdown or CSV feedback and flags configured phrases associated with frustration, wishes, switching and price. It groups matched text by configured themes and returns quotes, rule-match counts and matched-statement counts. One statement may match multiple rules; repeated text remains repeated occurrences.

    What you get

    A standard-library Python scanner, editable cue configuration, agent workflow, interpretation checklist and synthetic regression tests. JSON includes source paths and files-scanned coverage. Markdown presents grouped example quotes. These outputs help organize review; they do not establish genuine complaints, unique respondents, buying intent or demand.

    Select the right input

    For survey exports, identify the exact feedback headers and use --csv-column for each. Headers and unselected metadata are excluded. Without column selection, CSV is treated as headerless feedback-only text. UTF-8 and BOM-marked UTF-16 are supported. Missing or unreadable inputs, malformed or oversized CSV fields and folders with no supported files return no assessment. Linked and reparse-point inputs are refused.

    Boundaries

    Requires Python 3.10+ and an agent with authorized local execution. No scraping, network access or input-file writes. Matching is heuristic: negation, substrings, paraphrases and punctuation can cause false positives or misses. Scores count configured intensity words, not business priority. Confirm each quote before suggesting hypotheses or next steps. Price and existing terms remain unchanged.

    Find more skills like this on the Claude Skills Marketplace, the Codex Skills Marketplace and the Gemini Skills Marketplace, or browse Claude Skills for Research & Analysis.

    Keyword matching is not semantic complaint detection. Negations, substrings and punctuation can produce false positives or misses. Counts are text occurrences/rule matches, not people or demand. Intensity scores are keyword weights. Select CSV feedback columns explicitly; the header must be the first line. Duplicate selections, missing columns and unsupported inputs are refused. Generic errors require checking input format/permissions. Extreme directory depth, resource exhaustion or interrupted output is not a completed assessment. Windows link tests were limited by local privileges; independent link/permission tests ran on Linux.

    How to use

    1. 1

      Connect your agent to the Agensi MCP

      Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup

    2. 2

      Sign in to Agensi

      Your agent shows a sign-in link. Sign in and click Allow.

    3. 3

      Ask for what you need

      Your agent loads this skill and uses it on your task. It always asks before buying a skill.

    Reviews

    No reviews yet

    Be one of the first to try it. Every listed skill goes through the trust checks below.

    Security scanned

    Checked by our 8-point scan before listing

    Fresh listing

    Recently published to Agensi

    Money-back guarantee

    Doesn't do what its page promises? You get your money back.

    Trust & safety

    Security scan

    Security scan passed 16 days ago

    Software
    Python 3.10+Agent host with local terminal executionAuthorized feedback files

    Permissions

    Terminal / Shell
    Read Files

    File Scopes

    **/*.txt
    **/*.md
    **/*.csv
    --stdin

    Reads authorized local feedback plus bundled cue configuration. No network, credentials, environment-variable access or input-file writes. Invalid or unreadable input aborts assessment; linked/reparse inputs are refused.

    Listed5 months ago
    Updated16 days ago

    What's inside

    Frequently Asked Questions

    It provides a repeatable local keyword scan, explicit count definitions and traceable quotes for review. It does not establish an advantage for semantic interpretation, measured time savings or market validation. An agent may help interpret the results, but must verify context.

    Use a SKILL.md-capable agent that can execute Python 3.10+ against authorized local files. Synthetic tests ran locally on Windows and independently in Claude on Linux; those do not establish compatibility with every model or host.

    Use local .txt, .md or .csv files, folders, or plain feedback via stdin. CSV exports need explicitly selected feedback headers. UTF-8 and BOM-marked UTF-16 are supported. The helper does not scrape Reddit, review sites or support systems.

    The package contains SKILL.md instructions, a Python scanner, cue configuration, supporting review guidance, invented fixtures and regression tests. Existing price and terms are unchanged.

    The bundled references/pain-cues.json defines phrase rules, theme keywords and intensity words. Review edits and rerun synthetic checks; adding a keyword does not establish semantic accuracy or market evidence.

    Load the complete skill through Agensi into an agent with local Python execution. Provide authorized feedback and identify its feedback columns. A host that can only read instructions cannot execute the helper; technical assistance may be needed.

    New in Research & Analysis