Opportunities › Control plane for AI agent operations › Proposed app
Helmlog
One place to dispatch, watch and cost-track every AI agent your team runs
- Score
- 58
- confidence 49
- Evidence
- 5
- items on the page
- Rivals
- 1
- named in sources
- First version
- 40 d
- estimate
- Type
- Desktop and mobile
- PWA: on computers and phones alike
- Market since
- 2025
- New market · AI estimate
ScreensConcept
Desktop and mobile: Teams dispatch and configure agents at a desk, but need to stop runs and approve from phones while away from their setup. Sample screens on a computer and a phone, in the proposed brand. Their numbers are sample data.
Fleet
1 needs approval4 running
About this appAI proposal
Helmlog is a self-hosted control plane for teams that run more than one AI agent runtime. Connect Claude Code, Codex or another runtime through a small runner, then send tasks, follow each run and see what it cost, all from one screen.
It is for developers and small teams who host their own agents or mix self-hosted and hosted ones. Runners and run logs stay on your machines. The web app is a PWA, so you can check or stop a run from your phone.
Unlike a single-vendor dashboard, Helmlog treats runtimes as interchangeable: one task queue, one run history and one spend view, whichever agent did the work.
- Tagline
- Steer every agent from one wheel.
- Voice
- PlainSteadyPrecise
- Why this name
- A helm steers a ship and a log records the voyage. Together they say: control the agents and keep the record. Two plain syllables, easy to say.
In the first version
Members say which features they would use. It tells PWAS what people want; it changes no score.
Runtime-neutral task dispatch
Write a task once and send it to any connected runtime. Pick a target by name or label, and rerun the same task on another agent.
No member yet
Run review timeline
Open any run to see steps, tool calls, output and exit status in order. Mark runs accepted or rejected and leave a note.
No member yet
Spend tracking per run
Record API cost by run, runtime, project and person. Set a daily cap per runtime that pauses new dispatches when reached.
No member yet
Self-hosted runners
A light runner on your own machine or server pulls tasks and reports back. No inbound ports needed.
No member yet
Brand identityAI proposal
Check the name against trademark registers before using it.
Palette
#12363FDeep Harbor · primary#F5F1E8Chart Paper · surface#1B2228Ink · text#E8A33DSignal Amber · accent#3E8E7EKelp · data#C4472FBuoy Red · alert
Typography
Aa
Helmlog
PricingAI proposal
The material gives no competitor prices. These tiers are a starting point to test, based on the idea of pricing by runtime count and retention.
Community
Free
- Self-hosted
- 2 runtimes
- 7-day run retention
- Spend view
Team
$49 a month
- Up to 10 runtimes
- Roles and audit trail
- 90-day retention
- Spend caps and alerts
- Phone push alerts
Business
$199 a month
- Unlimited runtimes
- 1-year retention
- SSO
- Log export
- Priority support
What it could be worthEstimate
What a software business making this much sells for today, from real deals. Pick a monthly revenue.
If it makes
Worth about
$194,000
Sale range $132,000 to $263,000
$120,000 a year in revenue
From 87 software businesses listed and sold on Empire Flippers in the last year, revenue vetted by the broker: they sell for about 19× their monthly revenue, most between 13× and 26×.
Measured on profit, the middle deal is 32× monthly net profit: about 2.7 years of profit.
An estimate, not an offer or financial advice. A real price depends on profit, growth, churn, how much of the work is the owner's, and the buyer.
How it would growAI proposal
How the product would reach new customers through its own use, and how the first ones would be found. A plan to test, not a result.
- Built-in loop
- No loop in the product. Helmlog reaches customers through developer content, open docs and community channels. Run reviews stay inside teams. Optional shareable run links are a later test.
- Do they become customers?
- Test whether developers who self-host agents will pay for a managed or supported version, and whether a free self-hosted tier turns into Team plans.
- First customers
- Small engineering teams of 3 to 15 developers who already run both Claude Code and Codex on their own servers and have no shared view of spend.
- Where to reach them
- GitHub and open-source forums: Join discussions around self-hosted agent control planes and offer a free runner with a setup guide.
- Developer communities: Post spend-tracking write-ups and demos in agent and AI-coding communities.
- Direct outreach: Message engineering leads who publicly share agent workflows and offer a guided setup.
- First milestone
- Ten teams run Helmlog for two weeks and five of them pay for the Team plan.
Why people already pay for this
- VerifiedGitHub repository gained 6323 stars in 237 days
- Public claimSelf-hosted control plane for AI agents: dispatch tasks, review runs, track spend, and operate OpenClaw, Claude Code, Codex, and other runtimes
- VerifiedGitHub repository gained 977 stars in 29 days
- Public claimThe product provides a windowed desktop, isolated workspaces with VS Code, Claude Code and Codex, a store of self-hosted applications, and secure access via Tailscale only.
- Public claim
Tech stack
Every proposal builds on the same PWA base, installable from the browser on any phone or computer. Dashed boxes are what this app adds.
- AppNext.js and TypeScript
- PWAManifest and service worker
- UITailwind CSS
- Data and authSupabase Postgres
- PaymentsStripe
- EmailResend
- HostingVercel
- This app · Runner agentSmall cross-platform CLI daemonRuns on the customer's machine, pulls tasks and streams run events and cost back over outbound HTTPS.
- This app · Live run streamingServer-sent events or websocket serviceRun timelines and stop commands need near real-time updates.
RisksAI inference
- TechnicalEach runtime reports steps and cost differently. Adapters for Claude Code, Codex and others will break as those tools change.
- CommercialThe open-source project in the evidence is free and popular. Buyers may stay on it, so paid value must come from managed hosting, support or retention.
- PlatformRuntime vendors may add their own team dashboards and reduce the need for a neutral layer.
- DataRun logs can contain code and secrets. Storage, redaction and retention need care, or teams will not connect runners.
- AISpend figures depend on usage data from runtimes. If it is missing or late, caps may fail to stop overspend.
Everything marked AI proposal or AI inference was written by a model from the evidence on the opportunity page. The proposed name must be checked against trademark registers before it is used.