Partner toolkit · Reddit Lead Agent

Turn public Reddit threads into qualified signups.

A human-in-the-loop kit for finding public Reddit threads that match a SaaS offer, generating first-touch and follow-up reply packets, and recording every manually posted comment with proof. The same kit Comstein's founder uses to drive new SaaS signups from Reddit.

Files

Size

28 KB zip

Setup

~5 min

Runtime

Python stdlib

Apply as partner See the file tree

In production: Currently driving organic Reddit traffic and SaaS signups for the Comstein founder's own product.

queue/manual-ready-v2-cv-tool.md

# Manual Review Packet

- Subreddit: example
- Thread: Best tool to tailor my CV to job descriptions?
- Thread URL: https://www.reddit.com/r/example/comments/<thread_id>/slug/
- Saved: 2026-05-04T14:18:00Z
- Status: ready-for-manual-review
- Packet version: 2
- Draft validation: passed

## Why It Fits

The poster is looking for a tool that adapts a CV to specific job
descriptions with ATS-aware optimization — direct fit for the product.

## Source Notes
…

How it works

Six steps from public thread to recorded comment.

The kit drafts and organizes the work. You stay in control — every packet is reviewed and posted manually from your own browser profile, then the agent records public proof.

01
Discover
Pull public threads from configured subreddits and search queries.
02
Dedupe
Skip threads already in your manual-action ledger.
03
Draft
Generate first-touch and follow-up review packets.
04
Review
Edit the draft and validate against safety rules.
05
Post
Open in your own browser profile and post manually.
06
Record proof
Save the public comment URL to the ledger and verify it.

Sample outputs

The agent produces files you can review in any editor.

Every status update, queue entry, and review packet is a Markdown or JSON file. No hidden state, no platform lock-in. Switch tabs to see what each one looks like.

Examples shown — your own runs use your product, subreddits, and copy.

A status snapshot the build_dashboard_files.py script generates from the ledger.

# Latest Reddit Agent Status

- Generated: 2026-05-04 14:22 UTC
- Ready review packets: 4
- Follow-up packets: 2
- Posted comments: 12

Who uses it

Built for partners who actually run distribution.

SaaS founders

You ship your own SaaS and want a credible Reddit presence without hiring an agency or burning hours per week on outreach.

Niche creators

You promote tools to a focused audience and need a workflow that respects subreddit rules and your editorial voice.

Growth agencies

You run outbound for multiple SaaS clients and want one reusable workflow with proof recording and per-client config.

Outcomes

What this can produce — and what it costs to run.

The funnel below is a working model based on early Comstein partner runs. The token estimates assume you wire the kit to a Claude model for drafting, around 3 K input + 1 K output tokens per packet.

Per 100 threads discovered

Modeled weekly funnel

Pipeline

Threads discovered100

Packets ready for review30

Manually posted20

Resulting outcomes

Link clicks80

Signups6

AI cost per 100 generated packets

Token economics by model

Claude Haiku 4.5$1 /100

Best for high-volume drafting.

≈ $0.01 per packet

Claude Sonnet 4.6$5 /100

Balanced quality and cost.

≈ $0.05 per packet

Claude Opus 4.7$20 /100

Highest quality drafts.

≈ $0.20 per packet

Estimates only — not a guarantee. Actual numbers depend on your offer, audience, and how aggressive your approval rate is.

Setup

Setup time: ~5 minutes.

Pure Python, standard library only. No pip install, no Docker container, no platform lock-in. If you can run a Python script, you can run this kit.

1
Apply as partner
Submit the partner profile — the toolkit unlocks instantly in your dashboard.
2
Download and unzip
Open the Toolkit page in your partner workspace and download comstein-reddit-agent.zip.

Copy the templates

Set up the env file and config from the included examples.

$ cp .env.example .env
$ cp config/reddit-agent.config.example.json config/reddit-agent.config.json

4
Run the smoke check
Validate that everything is wired up before you discover your first thread.
```
$ python3 scripts/smoke_check.py --root .
```

Kit contents

What ships in the zip.

The kit is open and inspectable — no compiled binaries, no secrets, no telemetry. Sanitized of all live workspace data before bundling.

Package

Files: 39
Size: 28 KB zipped
Runtime: Python 3.10+ (standard library only)
Download: comstein-reddit-agent.zip

What you bring

Python 3.10 or newer.
Your own Reddit account.
A browser profile you control for manual posting.
Product, company, CTA, subreddit, and search-query configuration.

Manifest

reddit-agent-template/
├── README.md
├── START_HERE.md
├── ALLOWLIST.md
├── MARKETPLACE-LISTING.md
├── .env.example
├── config/  7 files — product config example, targeting + diversity rules, safety guidelines
├── docs/  5 files — install, operations, manual approval, dashboard, privacy
├── scripts/  12 files — 11 Python scripts (stdlib only) plus a scripts README
├── queue/  7 files — empty queue templates (Markdown + JSON)
├── ledgers/  1 file — manual-action ledger schema
├── reports/  2 files — report and comment-log templates
└── state/runner/  1 file — execution proof schema

What you get

Subreddit and search-query targeting tuned to your audience.
Deduplicated against a manual-action ledger so you never repeat replies.
First-touch and follow-up reply packets with safety rules baked in.
Proof recording for every comment URL you post manually.
Markdown and JSON outputs that work with any review surface.

Not in the box

No Reddit account, cookies, or auth files.
No browser profile or automated login.
No automatic posting bot — every reply is reviewed by you.
No live ledgers, queue packets, or customer data.

More kits

Coming next.

New channel kits land as we prove them on our own products before passing them to partners.

Kit	Channel	Status
`linkedin-outbound` Find founders, operators, and ICP personas where a partner has a credible warm-intro angle.	LinkedIn	Coming soon
`newsletter-pitch` Match SaaS offers to newsletter audiences and draft pitch-ready summaries with conversion-friendly framing.	Newsletter & Substack	Coming soon

Ready to plug the kit into your distribution?

Apply as a partner. The toolkit unlocks instantly after you submit your partner profile — no review queue.

Apply as partner