Autogrowth compiles your product listing once, then publishes it to software directories with formal, repeatable playbooks. Every submission is tracked, verified, and resubmitted on schedule. You do not fill forms by hand.
Every interaction with every directory is a written contract the engine executes the same way every time.
Enter your product once. Autogrowth builds the canonical listing payload: names, taglines and descriptions at every length budget directories ask for, pricing, media, and categories.
A validated playbook per directory drives a real browser session: sign in with email and password, fill declared fields, pass anti-spam checks, submit, and verify the result.
Every run lands in an append-only ledger with a formal state. Review-period directories are re-checked on schedule, and stalled or rejected listings are retried.
Each directory is a validated YAML contract: selectors, field bindings, anti-spam handling, guards, and success verification. Adding a directory is a data file, not new code, so coverage grows without regressions.
Re-running a submission can never double-submit. Terminal states short-circuit; retries resume safely.
Wire Autogrowth into Claude, Cursor, or any MCP client. Your agent can list directories, compile kits, preview submissions, and publish, all through typed tools.
Directories refresh, listings go stale, review queues back up. Autogrowth re-checks your listings and re-submits on a cadence, so visibility compounds instead of decaying.
Email and password signups are generated, stored, and reused per directory. No OAuth popups, no personal accounts, no browser-profile fragility.
The catalog is ordered by domain rating, traffic, tier, and automation readiness, so your first submissions always land where they matter most. Human-only platforms get prepared payloads instead of automation, respecting their terms.
Every run records steps, guards, screenshots, and final state. When a directory review asks what you submitted and when, the answer is one query away.
As of October 2026, typical directory tools stop at guided copy-paste. Autogrowth executes the submission.
| Capability | Guided checklists | Browser extensions | Autogrowth |
|---|---|---|---|
| Fills forms for you | Copy-paste | Autofill | Full submission run |
| Verifies the result | Manual | No | Declared checks |
| Idempotent retries | No | No | Ledger-backed |
| Agent access (MCP) | No | No | Native |
| Ongoing resubmission | No | No | Scheduled cycles |
| Entry price | $99/mo Pro tiers | Free-$29/mo | Free, then $19/mo |
Start free, keep the autopilot running for less than the cost of one shared lead.
It submits for you. For directories with validated playbooks, the runner opens a real browser session, authenticates with your vaulted email and password, fills the form, and submits. For platforms whose terms prohibit automation, Autogrowth prepares a complete submission payload for you to paste instead.
The run stops cleanly, records a NEEDS_HUMAN state with a screenshot, and retries next cycle. Autogrowth never bypasses bot protections.
No. The append-only ledger is checked before every run, and terminal states like SUBMITTED or LIVE short-circuit re-execution.
Point your MCP client at uv run python -m apps.autogrowth.mcp_server. Your agent gets tools to list directories, compile kits, preview and run submissions, and read the ledger, all rate-limited and audited by the same engine.
The catalog holds 1008 directories, including the major software directories, launch platforms, and AI tool directories. Playbook automation is live on 102 today and grows with each release; the rest follow the ranked queue.
Directory presence decays: listings get edited, sites restructure, review queues clear without notice. The subscription funds ongoing re-checks, resubmissions, and new playbooks, which is what keeps your visibility compounding.
Compile a kit, run the playbook, and see the ledger fill itself.
Start submitting