Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIMuseMuseDotsDots+21 more

    Hermes Cron Automation

    1

    Scheduled jobs for Hermes Agent that cost nothing when nothing happens and never fail silently — with tested watchdog scripts.

    $9

    Secure checkout via Stripe

    Money-back guarantee

    Converts to your local currency at checkout

    Security scanned
    Hermes Cron Automation

    Hermes Cron Automation

    Example session with this skill installed

    My Hermes jobs stopped delivering and I only noticed days later. Set up cheap watchdogs so this never happens again.

    • Read your context and instructions
    • Compiled the hermes cron automation

    What was wrong (diagnosed on a real Hermes install)

    hermes cron list showed four agent jobs failing

    13× in a row with HTTP 402: requires more credits — the OpenRouter balance was $0.00. The health check had warned, but nothing blocked it.

    Watchdogs added — all --no-agent, $0 per run

    cp scripts/{credit-watch.py,disk-watch.sh,service-watch.sh} ~/.hermes/scripts/   # copy, not symlink
    hermes cron create "0 */6 * * *" --no-agent --script credit-watch.py --deliver telegram --name credit-watch
    hermes cron create "every 30m"  --no-agent --script disk-watch.sh    --deliver telegram --name disk-watch
    hermes cron create "every 10m"  --no-agent --script service-watch.sh --deliver telegram --name gateway-watch
    

    Real output

    $ CREDIT_WATCH_MIN_USD=10 python3 credit-watch.py
    ⚠️ OpenRouter credit low: $5.00 left (threshold $10.00). Top up at https://openrouter.ai/settings/credits — agent jobs fail with HTTP 402 at $0.
    $ CREDIT_WATCH_MIN_USD=1 python3 credit-watch.py      # healthy → prints nothing → no message
    $ SERVICE_WATCH_USER=hermes-gateway bash service-watch.sh   # running → silent
    

    Change monitor instead of a 5-minute agent poll

    $ FEED_URL=https://github.com/NousResearch/hermes-agent/releases.atom FEED_MAX=3 python3 feed-monitor.py
    Hermes Agent v0.22.0 — The Threshold Release | tag:github.com,2008:Repository/1024554267/v0.22.0
    Hermes Agent v0.22.1 | tag:github.com,2008:Repository/1024554267/v0.22.1
    Hermes Agent v0.22.2 | tag:github.com,2008:Repository/1024554267/v0.22.2
    

    Used with --monitor-script feed-monitor.py, the model wakes only when this list changes.

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

    About this skill

    Automating recurring tasks in Hermes Agent often leads to silent failures or unexpected token drain. This skill provides the specific scripts and configurations needed to build high-reliability watchdogs and conditional agent jobs that only trigger when judgment is required.

    What it does

    • Automates watchdogs using script-only --no-agent modes to monitor disk usage, service status, and OpenRouter credits for $0 LLM cost.
    • Configures conditional triggers via --monitor-script and wakeAgent gates to prevent the model from waking up unless data has actually changed.
    • Suppresses unnecessary delivery by implementing [SILENT] logic in agent responses when no human intervention is needed.
    • Handles data refreshes through atomic write patterns and fallback chains that prevent corrupted upstreams from overwriting valid local snapshots.
    • Audits cron health using diagnostic commands to track execution history, incidents, and scheduler status.

    How it works

    1. Script deployment: Copy the provided .sh and .py monitoring scripts into the ~/.hermes/scripts/ directory.
    2. Job creation: Initialize a job using hermes cron create with the appropriate flags like --no-agent for simple checks or --monitor-script for LLM-driven analysis.
    3. Secret management: Configure local chmod 600 credential files for scripts to bypass the sanitized environment of the Hermes subprocess.
    4. Optimization: Pin specific models and minimize enabled_toolsets to reduce token overhead per execution tick.

    Frameworks & tools

    Targeted at Hermes Agent (Nous Research) v0.20.6 and above. Built using POSIX shell scripts, Python 3, and Hermes CLI commands.

    Why this beats prompting it yourself

    Prompting alone cannot solve the "silent failure" problem where a cron job stops running without alerting you. This skill provides the underlying system-level patterns, such as exit code handling and stable hashing for monitors, that generic prompting ignores.

    Use cases

    • Monitoring server health and API credits without incurring per-minute LLM costs.
    • Creating a daily morning briefing that only delivers if there are high-priority calendar events.
    • Building a robust RSS or API monitor that only alerts when new, specific keywords are detected.
    • Maintaining a local data snapshot with automated fallbacks and validity checks.

    Known limitations

    Scripts must be manually copied to ~/.hermes/scripts/ as symlinks are rejected. The scheduler cannot monitor its own host gateway status; external heartbeats are required for that.

    Scripts must be copied into ~/.hermes/scripts/ (symlinks pointing outside are rejected). The in-gateway scheduler cannot report that the gateway itself is down — use OS cron for that. Verified on Hermes v0.20.6; flag names can change between releases.

    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

    Requires

    Hermes Agent (v0.20+)
    Python 3.9+
    bash, curl
    Python
    Git

    Permissions

    Terminal / Shell
    Network Access
    Read Files
    Write Files
    Environment Variables

    Terminal: the scripts run as Hermes cron jobs. Network Access: credit-watch calls the OpenRouter credits API with your key; feed-monitor and safe-refresh fetch the URLs you configure. Environment Variables: thresholds and URLs. Read/Write Files: credentials file, state files and the refreshed snapshot.

    Listed1 month ago

    What's inside

    Frequently Asked Questions

    This skill provides system-level scripts and configurations that allow Hermes Agent to run background monitors for $0 LLM cost, only 'waking' the expensive model when specific conditions or changes are detected.

    It is specifically designed for Nous Research's Hermes Agent (v0.20.6+) and leverages the Hermes CLI and local file system patterns for high-reliability automation.

    Six tested scripts — disk-watch.sh, service-watch.sh, credit-watch.py, safe-refresh.sh, feed-monitor.py and wake-if-changed.sh — with 10 tests, the job-type decision guide and gotcha list in SKILL.md, and README.md.

    While basic prompting can describe a task, this skill provides the underlying exit-code handling, stable hashing, and atomic write patterns needed to prevent silent failures and data corruption in automated jobs.

    Setup involves moving scripts to your local Hermes directory and configuring a job via the CLI; however, please note that the host's gateway status requires an external heartbeat as the internal scheduler cannot monitor its own host connectivity.

    New in DevOps & Cloud