Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIMuseMuseDotsDots+21 more

    OHDA Professional troubleshooting

    1

    Professional technical troubleshooting using the OHDA method for replayable, structured worklogs and lessons learned lets your agents solve problems without going off the rails.

    $5

    Secure checkout via Stripe

    Money-back guarantee

    Converts to your local currency at checkout

    Security scanned
    OHDA Professional troubleshooting

    OHDA Professional troubleshooting

    Example session with this skill installed

    I need to debug why the staging database connection is timing out. Start an OHDA log.

    • Read your context and instructions
    • Compiled the ohda professional troubleshooting

    I have initialized a new OHDA worklog to track the database timeout issue. I will document the environment and initial observations before moving to hypotheses.

    ohda-2024-05-20-staging-db-timeout.md

    Objective: Restore connectivity to the staging PostgreSQL instance.
    Author: Claude Agent
    Date: 2024-05-20

    TL;DR: Connectivity failing with ETIMEDOUT. Investigating network ingress and credential validity.

    Reproduce:

    • Environment: K8s Namespace staging-apps
    • Target: postgres-staging.cluster.local:5432
    • Client: psql v15.3

    O1: Attempted connection from application pod api-v1-84fb.

    psql -h postgres-staging.cluster.local -U dev_user
    psql: error: connection to server at "postgres-staging.cluster.local" (10.96.0.10), port 5432 failed: Connection timed out
    

    H1: The SecurityGroup or NetworkPolicy is blocking ingress from the API pod.
    H2: The database service is down or not listening on the expected port.
    H3: The credentials in the db-secret have been rotated and are invalid.

    D1: kubectl get pods to check DB health (tests H2).
    D2: kubectl describe networkpolicy in the staging namespace (tests H1).

    Next steps

    1. Execute D1 to verify the database pod status.
    2. Check if the connection timeout occurs from other pods in the same namespace.
    ohda-2024-05-20-staging-db-timeout.md

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

    About this skill

    Stop losing track of complex debugging sessions and "why did I try that?" moments. This skill implements the OHDA method (Observe, Hypothesize, Decide, Act) to turn technical troubleshooting into a replayable, structured worklog. It forces a disciplined loop that prevents skipping straight to actions before documenting the reasoning and expected outcomes.

    What it does

    • Enforces OHDA loops by requiring a formal sequence of Observation, Hypothesis, Decision, and Action for every technical intervention.
    • Maintains replayable worklogs that document every command, timestamp, and raw output in an append-only Markdown format.
    • Manages external citations by pinning git commits, package hashes, and specific timestamps to ensure documentation stays valid as code evolves.
    • Tracks dead ends by striking through failed hypotheses rather than deleting them, preserving valuable "what not to do" data for the team.
    • Automates lesson promotion by summarizing closed-out logs and linking them in project-level metadata like CLAUDE.md.

    How it works

    1. Initialize a new log file with a unique topic name and the ohda-YYYY-MM-DD prefix.
    2. Execute loops by numbering steps (O1, H1, etc.) and documenting at least 2-3 hypotheses and candidate decisions before any mutation.
    3. Commit mutations by writing the A: (Act) step with rollback instructions before running the command, then closing the loop with a fresh O:.
    4. Close out the session by updating the TL;DR, performing a cold-reader check for local path leaks, and promoting the outcome to project memory.

    Frameworks & tools

    This tool operates via Markdown worklogs and integrates with local file systems and version control. It is designed to work within CLAUDE.md or similar project-context conventions.

    Why this beats prompting it yourself

    Generic troubleshooting prompts often lead to "fix-it" loops that disappear once the chat context clears. This skill enforces a specific IT industry standard for professional diagnosis that results in a permanent, searchable, and replayable asset for your entire team.

    Use cases

    • Deep-dive debugging where the root cause is non-obvious and requires multiple experiments.
    • Incident post-mortems that need a step-by-step audit trail of what was changed and why.
    • Handoffs and collaboration where a colleague needs to pick up a mid-diagnosis session without missing context.

    Known limitations

    Does not handle trivial one-shot fixes like typos. Requires a persistent file system to store logs and scripts.

    Not for trivial one-shot fixes. Requires manual confirmation for destructive actions. Log files can grow large if not closed out periodically.

    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 any payment.

    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 1 month ago

    Permissions

    Read Files
    Write Files

    No permissions needed. But the more access you give it, the more it can run automatically. Start by allowing it to observe the status of your problem state.

    Listed1 month ago

    What's inside

    Frequently Asked Questions

    The OHDA skill solves the 'context loss' problem in complex debugging by enforcing a structured loop of Observation, Hypothesis, Decision, and Action, ensuring every experimental step is documented and reversible.

    This skill is compatible with any AI agent that has file system access and supports Markdown, including Claude (via CLAUDE.md), ChatGPT, and local IDE-based agents.

    Upon purchase, you receive the OHDA implementation protocol, a library of Markdown log templates, and the automation instructions required to integrate the loop into your agent's workflow.

    Setup is simple: you provide the agent with the provided OHDA system instructions and designate a directory in your project for the worklogs to be stored and managed.

    While it excels at complex root-cause analysis, the OHDA skill is not recommended for trivial tasks like fixing typos or simple syntax errors where the overhead of a structured worklog outweighs the benefits.

    The skill includes automated 'lesson promotion' logic that summarizes the final log and updates your project’s global metadata, making the knowledge searchable for future sessions.

    New in Testing & Debugging