lukemcqueen/hermes-cortex - Page 4

shell-scripting

Shell scripting patterns, portability pitfalls, and cross-platform compatibility for bash/awk scripts in the Hermes Cortex ecosystem.

langfuse-self-hosted

Deploy, configure, and wire Langfuse v3 with ClickHouse for LLM observability — Docker compose, SIGSEGV-safe ClickHouse tuning, Hermes Langfuse plugin setup, OTLP endpoint troubleshooting.

cron-no-agent-conversion

Convert LLM-driven Hermes agent crons to no_agent scripts with targeted API calls. Maximizes deterministic Python work, uses LLM only for the single creative task the model uniquely provides.

llm-judge-scorer

LLM-as-Judge trace quality scorer. Evaluates Hermes conversation traces in Langfuse using a local Ollama model and posts quality scores back to Langfuse. Cron schedule: weekdays 12pm/8pm, weekends 10pm KST.

offline-code

Offline code snippet search + generation using local Ollama models. Search a 520-snippet corpus across 30+ topics and 25 languages with nomic-embed-text:v1.5, generate code with auto-detected qwen2.5 (3b on 4-8GB, 7b on 8-24GB, 14b on 24GB+). No internet…

cron-script-naming-reconciliation

Reconcile cron/script names, prefixes, skill versions.

cron-quality-gate

Prevents LLM cron jobs from delivering garbage with a self-check quality gate and automated watchdog.

prevent-crash-looping

How to prevent systemd service crash-looping from port conflicts, missing directories, and failed dependencies

orch-backlog-driver

Backlog-driven orchestrator work (F-023) — pull the top pending fleet tasks from the tasks DB, execute or dispatch them via the bus, verify with real evidence, close them, and report status to Telegram. The orchestrator's autonomous "nothing falls behind"…

task-persistence

Cross-session tasks via task-db.py or task_* MCP tools. v2: story/slices, paused, switch, bus-as-tasks, Telegram events.

cortex-deployment-sync

Use when pulling latest or running cortex update.

reflexion-check

Pre-delivery self-critique: seven-question audit to catch blind spots, verify claims, and score confidence before delivering results. Prevents half-baked work.

doc-freshness

Ensure AGENTS.md and SOUL.md stay current across all agents and projects. Weekly audit, post-update broadcast, mandatory section gap detection, and multi-project awareness. **Model requirement:** The audit cron needs 64K+ context window. Pinned to…

skill-miner

Mine loop governance DB, sessions, and memory for reusable skill patterns. Scores findings with nomic-embed-text and reports to the orchestrator via the agent inbox.

soul-refinement

Daily SOUL.md refinement process — mine sessions for lessons, apply corrections, codify principles. Optional per-agent Bible insight integration for daily Scripture reading.

nginx-security-pipeline

Set up nginx security with IP blocking, fail2ban integration, daily automated scanning, and atomic deploy. Covers blocked_ips.add input, fail2ban filter, deploy script (backup → validate → reload), and a no_agent daily scanner cron. Platform-aware: macOS…

threat-defense-pipeline

Layered defense system: fail2ban jails + nginx IP blocking + daily auto-deploy pipeline. How to add blocked IPs, check jails, and deploy.

cron-output-contracts

Use when a script reads another cron's output or retires.

eval-harness

Systematic evaluation framework for agent capabilities — capability tests, regression suites, failure analysis

two-hard-rules

Two hard rules every agent must follow: USE LOOP GOVERNANCE ALWAYS. SHARE IMPROVEMENTS TO THE PUBLIC REPO.

migration-deploy-verification

Use when a repo migration is missing from the live DB.

pgmq-consumer-patterns

Use when fixing bus consumers or false-unreachable agents.

telegram-delivery-diagnostics

Diagnose and fix Telegram delivery issues for Hermes cron jobs — delivery pipeline tracing, DNS/network diagnostics, adapter state verification, and known failure patterns.

architecture-review

Multi-role architecture review (a.k.a. HC-Party) with weighted decision matrices, conflict resolution, and cost estimation.

requirements-elicitation

Requirements elicitation for Hermes Cortex (a.k.a. elicit) — structured domain exploration, RICE/MoSCoW prioritization, and user-story output with acceptance criteria.

agent-fundamentals

Universal 'basic things every agent should know' — distilled from real frustration patterns across 10+ sessions. Prevents the most common corrections and wasted turns.

sudoers-audit

Systematically audit, test, and debug sudoers NOPASSWD rules — verify allowed commands actually run without password, diagnose silent rule failures, and fix common pitfalls.

task-decomposition

Break large tasks into discrete, verifiable, independently completable units. Uses functional decomposition, vertical slicing, dependency ordering, and scope gating so agents execute predictably end-to-end.

hermes-desktop-plugins

Write desktop app plugins that add UI panes and commands.

hermes-themes

Author a Hermes color theme that skins every surface.

memory-architecture

Design and maintain agent memory system: MEMORY.md structure, privacy boundaries, gitignore per brain source, seed templates, pointer pattern, and PII prevention. Use when asked: 'clean up memory', 'separate public from private', 'prevent PII leakage',…

deployed-component-verification

Verify deployed components match their repo source — detect stale copies, validate symlinks, and ensure the fleet runs fresh code after git pulls.

cold-email

Write B2B cold emails and follow-up sequences that get replies. Use when the user wants to write cold outreach emails, prospecting emails, cold email campaigns, sales development emails, or SDR emails. Also use when the user mentions "cold outreach,"…

copywriting

When the user wants to write, rewrite, or improve marketing copy for any page — including homepage, landing pages, pricing pages, feature pages, about pages, or product pages. Also use when the user says "write copy for," "improve this copy," "rewrite this…

marketing-psychology

When the user wants to apply psychological principles, mental models, or behavioral science to marketing. Also use when the user mentions 'psychology,' 'mental models,' 'cognitive bias,' 'persuasion,' 'behavioral science,' 'why people buy,' 'decision-making,'…

programmatic-seo

When the user wants to create SEO-driven pages at scale using templates and data. Also use when the user mentions "programmatic SEO," "template pages," "pages at scale," "directory pages," "location pages," "[keyword] + [city] pages," "comparison pages,"…

email-sequence

Write a multi-email nurture/onboarding/launch sequence with a goal per email. Use when asked to write an email sequence, a welcome/onboarding series, a nurture drip, a launch sequence, or a re-engagement series. Produces the sequence map (trigger, timing,…

social-content

When the user wants help creating, scheduling, or optimizing social media content for LinkedIn, Twitter/X, Instagram, TikTok, Facebook, or other platforms. Also use when the user mentions 'LinkedIn post,' 'Twitter thread,' 'social media,' 'content calendar,'…

app-store-optimization

App Store Optimization (ASO) toolkit for researching keywords, analyzing competitor rankings, generating metadata suggestions, and improving app visibility on Apple App Store and Google Play Store. Use when the user asks about ASO, app store rankings, app…

brand-guidelines

When the user wants to apply, document, or enforce brand guidelines for any product or company. Also use when the user mentions 'brand guidelines,' 'brand colors,' 'typography,' 'logo usage,' 'brand voice,' 'visual identity,' 'tone of voice,' 'brand…