My Patreon Manager

Automate tier-gated content creation for Patreon from your Git repositories.

My Patreon Manager scans repositories across GitHub, GitLab, GitFlic, and GitVerse, generates quality-scored content through an LLM pipeline, and publishes tier-gated posts to Patreon. CLI-first, idempotent, and safe to re-run.

---

---

Find What You Need (Landing)

Jump directly to the information you're looking for:

Setting up for the first time

• Quick Start (below) — 5 steps from clone to publish
• Quickstart Guide (long-form) — the full 10-minute walkthrough
• Obtaining Credentials — step-by-step signup flows for every token-issuing service with direct links (Patreon · GitHub · GitLab · GitFlic · GitVerse · OpenAI DALL-E · Stability AI · Midjourney proxies · OpenAI-compatible providers · LLMsVerifier)
• Configuration Reference → Required / Optional at a Glance — every env var grouped by flow with mandatory/optional markers and direct signup links
• Minimum viable
  .env
  recipe — the smallest config that delivers multi-org scanning + articles + illustrations

Running the pipeline

• process
  command design — the versioned, revision-aware pipeline that replaces
  sync
• process
  implementation plan — task-by-task walkthrough of what's built and how
• Configuration Reference → Process Pipeline — every
  MAX_ARTICLES_*
  ,
  GENERATOR_VERSION
  , drift-check, and lock knob
• CLI: process · scan · generate · publish · validate

Migrations and schema

• Migration system refactor spec — why the versioned Migrator exists
• patreon-manager migrate up | status | down
  — CLI for schema migrations
• SQL Schema reference

Articles + illustrations

• Content Generation Guide — LLM pipeline, quality gates, tier mapping
• Illustration Generation + Image Providers — DALL-E 3 · Stability AI · Midjourney · OpenAI-compat
• LLMsVerifier Setup — the quality-gate service all LLM calls route through

Patreon specifics

• Patreon Tiers Guide — tier configuration and access control
• Obtaining Credentials § Patreon — OAuth walkthrough + campaign-ID lookup

Deployment, operations, security

• Deployment Guide · Docker · Podman · systemd · Kubernetes · Bare Binary
• Admin Manual — webhooks, monitoring, SLOs, credential rotation
• Security docs · Runbooks · Troubleshooting FAQ
• Local Verification (15-step pre-publish checklist)

Architecture and development

• Architecture Overview · ADRs
• Developer Manual — adding providers, renderers, migrations, tests
• Tutorial — first sync · server setup · security scanning · testing · content pipeline
• OpenAPI spec · CLI reference
• Main specification

Project status and future work

• Known Issues, Unfinished Work & Future Enhancements — the canonical "what's not done and why" document. Covers product non-goals (multi-tenancy, drift auto-merge, article scheduling, separate REVIEWER_KEY), infrastructure gaps (Postgres integration harness, coverage.sh measurement semantics, submodule mirrors), deferred enhancements (
  models.Post.URL
  , illustration cleanup on repo delete, preview UI SPA, webhook-driven sync, multi-node parallelism, migrate-down backups), documentation deferrals (video course scripts, legacy planning artifacts), and environmental caveats (Semgrep auth, COVERAGE_MIN default). Each entry has a workaround and a concrete path to resolution.

Full documentation index with every file: see "Documentation" section below.

Features

• Multi-platform Git scanning -- GitHub, GitLab, GitFlic, and GitVerse as first-class, interchangeable sources with mirror detection
• Multi-organization support -- scan repositories across multiple organizations and groups in a single sync run
• LLM-powered content generation -- quality-scored model selection with automatic fallback chains and configurable quality thresholds
• Tier-gated Patreon publishing -- maps repository content to Patreon tiers with deduplication via content fingerprinting
• CLI subcommands --
  process
  (top-level pipeline),
  scan
  ,
  generate
  ,
  validate
  ,
  publish
  with
  --dry-run
  ,
  --schedule
  ,
  --org
  ,
  --repo
  ,
  --pattern
  ,
  --json
  ,
  --log-level
  .
  sync
  is retained as a deprecated alias for
  process
  .
• HTTP server -- Gin-based server on
  :8080
  with health checks, Prometheus metrics, admin endpoints, and webhook handlers
• Resilience patterns -- circuit breakers, exponential backoff, per-provider rate limiting
• Observability -- structured JSON logging, Prometheus metrics, Grafana dashboards
• Idempotent operations -- content fingerprinting and checkpointing ensure safe re-runs after failures
• Database flexibility -- SQLite (default) or PostgreSQL for production
• Security-first -- twelve-factor credential management, HMAC-signed webhooks, credential redaction in all logs

Quick Start

sync

is still accepted as a deprecated alias for

process

and prints a warning to stderr. Prefer

process

for new automation.

Environment Variables Quick Reference

Full reference with mandatory/optional tables by flow: docs/guides/configuration.md § Required / Optional at a Glance Step-by-step signup flows with direct links to every token-issuing service: docs/guides/obtaining-credentials.md

Required

Legend: ❗ mandatory · ○ optional.

Patreon API (all ❗)

Security

Recommended

Git Provider Tokens (at least one ❗ for multi-org scanning)

Each provider also supports a

token for failover (e.g.,

GITHUB_TOKEN_SECONDARY

).

Multi-Organization (all ○)

Illustration Providers (set at least one ❗ to generate per-article images)

Illustration behavior is controlled by

(default

true

),

IMAGE_PROVIDER_PRIORITY

(default

dalle,stability,midjourney,openai_compat

),

ILLUSTRATION_DEFAULT_STYLE

,

ILLUSTRATION_DEFAULT_SIZE

,

ILLUSTRATION_DEFAULT_QUALITY

, and

ILLUSTRATION_DIR

. See Illustration Generation and Image Providers for full reference.

LLMsVerifier (required for content generation)

Optional

Content Generation

Database

Server

Multi-Org Configuration

Configure multi-organization scanning to discover repositories across multiple GitHub organizations and GitLab groups in a single sync run:

When unset, each provider scans the token owner's personal repositories. All organizations are iterated with full pagination, and cross-org deduplication is applied through the existing mirror detection pipeline.

See Multi-Org Support for complete documentation.

Documentation

Getting Started

Architecture

Guides

API Reference

Tutorials

Manuals

Video Course (11 modules)

See also: Course Outline | Recording Checklist | Distribution Plan

Operations

Development

License

See the project license file for terms and conditions.