Grant
COO
Answer your questions about your software operations, send each piece of work to the teammates who own it, and follow your decisions.
His team’s briefs, the findings and your decisions
What Grant doesHire Grant. Your COO and his ops team.
Grant and his specialists watch your CI, cloud costs, runtime, security and dependencies, then bring you what matters with the evidence in Slack, the dashboard or MCP. They’re all AI agents, and nothing merges or deploys.
Grant’s team
Grant
David
Maya
Liz
Noah
Priya
Owen
Iris
Harper
Theo
Vera
Since Oct 8: 37 public changesReduced motion
Since : 37 public changes across crates.io, GitHub and npm.
Hire Grant, get the whole team behind him.
Grant leads AI specialists for CI, cloud costs, runtime, security and dependencies, and brings you answers with the evidence attached.
Illustrative activity Every label is one the product shows while the team works.
David
Compatibility and repairs
Find what upstream changes, deprecations and end-of-life dates mean for the company’s code, and report the repair pull requests that fix them.
World State changes · Repository context · GitHub delivery
What David doesMaya
CI and delivery
Find why CI is slow or failing, and say in plain words what shipped and how fast the team delivers.
GitHub Actions · Repository context · GitHub delivery
What Maya doesLiz
Costs
Track what the company spends on AWS and AI providers, where it is heading against each budget, and what is paid for but unused.
AWS costs · AWS forecast and budgets · GitHub seats · AI provider spend · AWS commitments · AWS running resources · GitHub delivery · AWS changes
What Liz doesNoah
Runtime
Watch the health of what the company runs in AWS, and say what changed when something breaks.
AWS runtime metrics · AWS inventory · AWS changes · GitHub delivery · AWS quotas and limits · AWS running resources
What Noah doesPriya
Security
Find the security issues that matter most across the company’s code, GitHub and AWS, and prepare the evidence reviewers and auditors ask for.
GitHub security alerts · AWS Security Hub · Repository context · GitHub access · AWS security posture · GitHub delivery · AWS changes · GitHub seats · AI provider members and keys · Slack users
What Priya doesOwen
Configuration and lifecycle
Keep track of how the company’s infrastructure is configured and of every date that is about to run out.
Repository context · World State changes · AWS inventory · AWS security posture
What Owen doesIris
World State
Keep watch on the outside world the company’s projects depend on, and check facts when asked.
World State changes
What Iris doesHarper
Chief of Staff
Turn the team’s saved briefs into documents an executive can read: reports, one-pagers, decks and the Monday brief.
Her teammates’ saved briefs, and the company documents a request names
What Harper doesTheo
Setup and standards
Find what the company has not set up yet in AWS and GitHub, and keep one setup plan, most important first.
AWS security posture · Repository context · AWS costs
What Theo doesVera
Agent security
Test the company’s own AI agents the way an attacker would, every week, and report what got through.
Agent endpoint
What Vera doesPut them to work
Ask Grant anything. Supported fixes arrive as draft PRs, and you set the project scope. Free answers from saved results; paid plans consult the team live.
Hire Grant free
Ask Grant. His team does the legwork.
Ask in plain words. Grant brings in the right specialists, shows their sources, and keeps long jobs going until they report back.
Grant asks up to six teammates at once and answers with citations. Paid plans do this live.
Work still running after about 45 seconds, or when you ask, keeps going in the background and reports back in the same conversation.
The team checks in on your plan’s schedule (daily on Pro, every six hours on Studio, and once a day for Liz and Owen) and when GitHub reports a workflow run, a check or a push on the default branch, or a new alert.
Assign an owner, defer to a date or mark it addressed. The next check-in picks up where you left off. Marking it addressed records your decision; it isn’t proof. A disconnected account holds up only its own sources.
GrantFollow-upNoah
Lambda errors on checkout-api went from 0.2% to 1.1% after Tuesday’s deploy.
In the same conversation
Check-insDaily on Pro
- Default-branch runs and checksMaya
- Default-branch pushesDavid and Owen
- New alertsPriya
Every six hours on Studio; Liz and Owen once a day
The test job is taking longer
Maya · CI and delivery
- Owner @sam
- Deferred
- Revisit on Friday
At the next check-in, new runs are compared with the saved finding.
Outside change. Inside fix.
Connect your GitHub repository. See what changed, find the affected code, and review a supported repair.
- API changes
- SDK updates
- Deprecations
- Provider changes
Illustrative workflow. Fictional package and observations.
Supported patches and draft PRs by Since.dev. Tests run in your repository.
Since.dev Draft PR
Small patch. Clear evidence.
Removed: client.sendAdded: client.messages.createDavid · Reports open repair PRs
Repairs that fit your workflow.
Since.dev traces supported upstream changes to affected code and prepares a focused patch.
Your checks, already in place.
Draft PRs use your existing CI.
Revisions on the same PR.
Request a revision in the dashboard, or mention the installed GitHub app in a comment on its repair PR. Since updates the same PR and reruns its checks.
The source behind the fix.
Upstream evidence stays with the repair.
One repair a month on Free.
Free repositories stay in Monitor only. The allowance is used only when a draft PR opens; failed or rejected attempts do not use it.
Ask where you already work. Same Grant. Same team. Same findings.
Ask in a Slack thread, the dashboard or your agent’s MCP client. Each one reaches the same Grant, his team and your findings.
Like a teammate in the thread. DMs are private and persistent. Nothing Grant says is ephemeral. Channel answers stay within the projects the connection covers.
Watch the team work. Stop any one of them.
Your agent can ask too. 21 tools; start with grant_team_status. Waits up to 55 seconds, then call again with the same requestKey.
Connect in a minute: MCP, an HTTP API and signed webhooks. Catch up on changes since your agent’s last acknowledged update. New watches cover future changes only.
Slack
Illustrativeops
Grant App
@sam asked Grant
how’s my AWS ops?
Grant App
@sam asked Grant
Done · consulted Liz, Noah Liz · Done Noah · Done
Grant App
Answer for @sam
Compute spend rose from $120 to $156 across same-length periods (Liz). Lambda errors on checkout-api went from 0.2% to 1.1% after Tuesday’s deploy (Noah).
/grant how’s my AWS ops?
Dashboard
IllustrativeHow’s my AWS ops?
Consulted Liz, Noah and Priya
- LizDone
- NoahDone
- PriyaStopped. Partial results are saved.
Grant
Compute spend rose from $120 to $156 across same-length periods (Liz). Priya was stopped, and her partial results are saved. Noah is still working and will post here when he’s done.
GrantFollow-upNoah
Lambda errors on checkout-api went from 0.2% to 1.1% after Tuesday’s deploy.
MCP
IllustrativeCall
grant_question({
input: { content: "How's my AWS ops?",
conversationId: null, waitSeconds: 55 },
requestKey: "ops-review-0929"
})After up to 55 seconds
{ "status": "running",
"answer": null,
"team": [{ "name": "Liz",
"activity": "AWS costs" }, … ] }The same call again
{ "status": "completed",
"answer": {
"content": "Compute spend rose…",
"references": [ … ] }, … }Iris · World State
Watch what your code depends on.
Get API changes, security fixes, and new releases for the things you watch.
Your watch instructions and notifications stay private.
How watches workIrisWorld StateIllustrative
A watch found a change
example-sdk 19.0.0 to 20.0.0breaking
client.send is replaced by client.messages.create.
Official release notes and migration guide
Give your agents a reality check.
Valid, changed or unknown against the latest recorded observation, with its observation time and source health. Unknown does not mean unchanged.
That’s world state: external facts, versions, and provider conditions, ready for your agents to check.
Explore verificationDoes this still hold?
example-sdk.version = 19.0.0Something changed
20.0.0The observation differs. Your agent gets evidence to review.
Illustrative results · check observation time and source health. Verification reads recorded evidence, without a fresh fetch.
A clearer view of what changed.
Browse published package and API changes, each linked to its evidence.
Explore changesvitejs/vite: v8.3.3
behavioralvitest-dev/vitest: v5.0.1 → v5.0.2
behavioralbetter-auth/better-auth: 1.7.6
behavioral
A little context for every change.
- breaking
- deprecated
- behavioral
- additive
- advisory
Start free. Add the team when you’re ready.
Free answers from saved results; paid plans consult the team live on Since AI, with your own AI provider as an option.
- No card required
- Daily check-ins on Pro
- AI included on every plan
Grant Free
Meet Grant with one project and a monthly operational baseline.
$0 / month
Free, with no card required
Start freeIncludes
- Setup: repository map
Grant Pro
Daily check-insDaily check-ins across five projects.
$49 / month
$40 a month billed yearly, 2 months free
Start freeAdds to Free
- Reports: company operations report, Monday ops brief
- Alerts and follow-up: alerts you ask for, owner reminders and accepted-risk reviews
- Money: forecast and budget variance, monthly bill explained, AWS credits runway, running and forgotten resources
- Security: security questionnaire drafts, vulnerabilities ranked by exploit risk
- Delivery: what shipped this week
- Lifecycle: end-of-life calendar, things about to expire, database and backup safety check
- Setup: top 5 setup plan items
Grant Studio
Check-ins every six hours across fifteen projects.
$149 / month
$124 a month billed yearly, 2 months free
Start freeAdds to Pro
- Reports: one-pagers for a named reader, board and investor decks
- Alerts and follow-up: incident follow-up tracking
- Money: AI and LLM spend, SaaS seats and waste, savings tracker, commitment coverage and expiry
- Security: access reviews, audit evidence packs, security alert fix deadlines, GitHub Actions supply-chain audit, change management evidence, offboarding check and leaver sweep, risky change tripwires
- Reliability: alarm first responder, monthly reliability review, deploy verification, database storage runway, quota and concurrency runway
- Delivery: delivery pace, incident postmortem drafts, CI health and toil report, review bottlenecks
- Lifecycle: vendor risk watch, vendor and subprocessor register, new dependency intake and licence report
- Setup: full setup plan
- Company knowledge: uploaded company documents, proposal reviews
- Agent security: weekly tests of your own agents
Grant Enterprise
Agreed project scope, check-in cadence and support for your team.
Custom
Terms agreed with your team
Contact usAdds to Studio
- Security: evidence across the audit period
Pro and Studio keep working past their credits: extra credits are billed at $2.00 per 1,000 on Pro and $1.50 per 1,000 on Studio, on a separate monthly invoice, with no cap. Free pauses Since AI until the 1st. Work on your own AI provider uses no credits.
See everything in each plan
Free includes Grant plus David, Maya, Liz and Priya in the dashboard, the last 30 days of history (older records are kept) and 1,000 Since AI credits a month. Pro adds Noah, Owen, Iris, Harper and Theo, Grant in Slack and MCP, full history and daily check-ins. Pro also includes the company operations report, Monday ops brief, alerts you ask for, owner reminders and accepted-risk reviews, forecast and budget variance, monthly bill explained, AWS credits runway, running and forgotten resources, security questionnaire drafts, vulnerabilities ranked by exploit risk, what shipped this week, end-of-life calendar, things about to expire, database and backup safety check and top 5 setup plan items. Studio adds Vera (agent security), one-pagers for a named reader, board and investor decks, incident follow-up tracking, AI and LLM spend, SaaS seats and waste, savings tracker, commitment coverage and expiry, access reviews, audit evidence packs, security alert fix deadlines, GitHub Actions supply-chain audit, change management evidence, offboarding check and leaver sweep, risky change tripwires, alarm first responder, monthly reliability review, deploy verification, database storage runway, quota and concurrency runway, delivery pace, incident postmortem drafts, CI health and toil report, review bottlenecks, vendor risk watch, vendor and subprocessor register, new dependency intake and licence report, full setup plan, uploaded company documents and proposal reviews. Paid plans consult the team live on Since AI. To use your own AI provider, connect Anthropic or OpenAI on Pro. Studio and Enterprise also support AWS Bedrock and any OpenAI-compatible endpoint, such as OpenRouter or Moonshot, by its HTTPS base URL. Any model ID the connection test marks ready for Grant works.
Grant Free
- 1 active project
- 1 baseline + 5 follow-up check-ins per UTC month
- 1,000 Since AI credits a month
- Grant plus David, Maya, Liz and Priya
- Grant in the dashboard only
- 30 days of history
- 50 repository dependencies monitored
- 5 World State watches, every aspect of each
- 1 draft repair PR a month, on demand
- 100 verification calls a month
- 1 workspace member
Includes
- Setup: repository map
Grant Pro
- 5 active projects
- Daily check-ins
- 25,000 Since AI credits a month, then $2.00 per 1,000 extra credits
- Grant and the whole team except Vera
- Grant in Slack, MCP and the dashboard
- Full history
- Optional: bring your own Anthropic or OpenAI key
- No repository dependency count limit
- 150 World State watches, every aspect of each
- Automatic repairs and draft PRs
- 500 verification calls a month
- 3 workspace members
Includes
- Reports: company operations report, Monday ops brief
- Alerts and follow-up: alerts you ask for, owner reminders and accepted-risk reviews
- Money: forecast and budget variance, monthly bill explained, AWS credits runway, running and forgotten resources
- Security: security questionnaire drafts, vulnerabilities ranked by exploit risk
- Delivery: what shipped this week
- Lifecycle: end-of-life calendar, things about to expire, database and backup safety check
- Setup: repository map, top 5 setup plan items
Grant Studio
- 15 active projects
- Check-ins every 6 hours
- 80,000 Since AI credits a month, then $1.50 per 1,000 extra credits
- Grant and the whole team
- Grant in Slack, MCP and the dashboard
- Full history
- Optional: bring Anthropic, OpenAI, AWS Bedrock or any OpenAI-compatible endpoint
- No repository dependency count limit
- 500 World State watches, every aspect of each
- Automatic repairs and draft PRs
- 2,000 verification calls a month
- 10 workspace members
Includes
- Reports: company operations report, one-pagers for a named reader, board and investor decks, Monday ops brief
- Alerts and follow-up: alerts you ask for, owner reminders and accepted-risk reviews, incident follow-up tracking
- Money: forecast and budget variance, AI and LLM spend, SaaS seats and waste, monthly bill explained, AWS credits runway, running and forgotten resources, savings tracker, commitment coverage and expiry
- Security: security questionnaire drafts, access reviews, audit evidence packs, vulnerabilities ranked by exploit risk, security alert fix deadlines, GitHub Actions supply-chain audit, change management evidence, offboarding check and leaver sweep, risky change tripwires
- Reliability: alarm first responder, monthly reliability review, deploy verification, database storage runway, quota and concurrency runway
- Delivery: what shipped this week, delivery pace, incident postmortem drafts, CI health and toil report, review bottlenecks
- Lifecycle: end-of-life calendar, things about to expire, database and backup safety check, vendor risk watch, vendor and subprocessor register, new dependency intake and licence report
- Setup: repository map, full setup plan
- Company knowledge: uploaded company documents, proposal reviews
- Agent security: weekly tests of your own agents
Grant Enterprise
- Custom project allowance and cadence
- Since AI credits set by contract
- Optional: bring Anthropic, OpenAI, AWS Bedrock or any OpenAI-compatible endpoint
- Workspace limits agreed by contract
- Sales-led onboarding and billing
Includes
- Reports: company operations report, one-pagers for a named reader, board and investor decks, Monday ops brief
- Alerts and follow-up: alerts you ask for, owner reminders and accepted-risk reviews, incident follow-up tracking
- Money: forecast and budget variance, AI and LLM spend, SaaS seats and waste, monthly bill explained, AWS credits runway, running and forgotten resources, savings tracker, commitment coverage and expiry
- Security: security questionnaire drafts, access reviews, audit evidence packs, vulnerabilities ranked by exploit risk, security alert fix deadlines, GitHub Actions supply-chain audit, change management evidence, offboarding check and leaver sweep, risky change tripwires, evidence across the audit period
- Reliability: alarm first responder, monthly reliability review, deploy verification, database storage runway, quota and concurrency runway
- Delivery: what shipped this week, delivery pace, incident postmortem drafts, CI health and toil report, review bottlenecks
- Lifecycle: end-of-life calendar, things about to expire, database and backup safety check, vendor risk watch, vendor and subprocessor register, new dependency intake and licence report
- Setup: repository map, full setup plan
- Company knowledge: uploaded company documents, proposal reviews
- Agent security: weekly tests of your own agents
Good questions. Clear answers.
Read the docsWho is Grant?
Grant is the COO for your software operations, an AI agent built by Since.dev. He leads a preset team of specialists who watch your GitHub Actions CI, cloud and AI costs, runtime metrics, existing GitHub and cloud security findings, repository configuration and lifecycle, and dependency compatibility. Each reads its own sources and saves a brief. On paid plans Harper turns their briefs into reports and the Monday ops brief, and Theo keeps a setup plan for your cloud account and repositories; on Studio and Enterprise, Vera also tests your own agents for prompt injection and other agent risks. Ask Grant in Slack, the dashboard or your agent's MCP client: he consults the teammates your question needs, answers in one voice with the evidence and keeps track of what you decide. Free answers in the dashboard from saved results; paid plans consult the team live on Since AI, or on your own AI provider if you choose.
Are Grant and his team real people?
No. Grant and every specialist on his team are AI agents built by Since.dev, not people. They read the sources you connect; supported repairs arrive as pull requests for your review, and nothing merges or deploys. Grant covers your software operations only: he doesn't hire, manage people or handle finance.
What's the difference between Free and paid?
Free covers one project: one baseline and five follow-up check-ins per UTC month. Free includes Grant plus David, Maya, Liz and Priya in the dashboard, the last 30 days of history (older records are kept) and 1,000 Since AI credits a month. Grant answers from saved results and says as of when; he does not consult the team live. Pro adds Noah, Owen, Iris, Harper and Theo, Grant in Slack and MCP, full history and daily check-ins. Pro also includes the company operations report, Monday ops brief, alerts you ask for, owner reminders and accepted-risk reviews, forecast and budget variance, monthly bill explained, AWS credits runway, running and forgotten resources, security questionnaire drafts, vulnerabilities ranked by exploit risk, what shipped this week, end-of-life calendar, things about to expire, database and backup safety check and top 5 setup plan items. Studio adds Vera (agent security), one-pagers for a named reader, board and investor decks, incident follow-up tracking, AI and LLM spend, SaaS seats and waste, savings tracker, commitment coverage and expiry, access reviews, audit evidence packs, security alert fix deadlines, GitHub Actions supply-chain audit, change management evidence, offboarding check and leaver sweep, risky change tripwires, alarm first responder, monthly reliability review, deploy verification, database storage runway, quota and concurrency runway, delivery pace, incident postmortem drafts, CI health and toil report, review bottlenecks, vendor risk watch, vendor and subprocessor register, new dependency intake and licence report, full setup plan, uploaded company documents and proposal reviews. Paid plans consult the team live in Slack, the dashboard and MCP, check in daily on Pro or every six hours on Studio (Liz and Owen once a day), and also check in when GitHub reports a workflow run, a check or a push on the default branch, or a new alert. Since AI is included on every plan, with 25,000 credits a month on Pro and 80,000 on Studio. Your own AI provider is optional on paid plans.
What do I need to connect?
Start on Home: choose GitHub or AWS and connect it. On GitHub, install the Since.dev app and pick a repository; it becomes your first project. On AWS, create a read-only role from the template we prepare, then name the project and choose what the team reads. Nothing else is needed to start: Since AI is included on every plan, and on a paid plan you can connect your own AI provider instead. Choose Approve and start check-in, and your team runs its first check-in right there. Only following public sources? Create a World State watch instead. Existing repositories and watches remain connected. Slack and signed webhooks are optional. Missing connections and unsupported resources appear as coverage gaps.
Can Grant change my cloud or production systems?
No. Check-ins and consults only read the sources you connect and the documents you upload, except Vera's tests of your own agents on Studio and Enterprise. Grant can record decisions and follow up, but cannot change cloud resources, rerun workflows or enable scanners. The existing repair flow can open draft pull requests for human review, and nothing merges on its own. Compatibility repairs keep their repository permissions, eligibility and repair settings. A cloud or security finding does not grant permission to write code or make production changes.
What can Grant see in Slack?
Once your Slack and Since identities are linked (automatically when your Slack email matches your Since account), ask with /grant or a mention in a channel: Grant answers publicly in that thread and discusses only the connection's projects, every project unless an owner narrowed them. DM Grant for a private, persistent chat over your own access. Grant reads only messages addressed to him (mentions, /grant and DMs with him), never broad channel history.
Does it repair my code automatically?
New paid repository connections default to Automatic draft PRs. Since automatically checks impact, prepares supported repairs, and opens draft PRs. Choose Monitor only to prepare repairs on demand. Repairs cover small, documented compatibility fixes in code and configuration. Routine version bumps and lockfile changes are outside that scope. If the evidence is insufficient, no PR is created. That does not prove your code is unaffected. Where GitHub doesn't offer draft PRs (private repositories on GitHub Free or Pro), the repair opens as a regular PR titled [Needs review]. CI runs in your repository. Since.dev never merges or deploys.
Can I use my own AI provider?
Yes, on Pro, Studio and Enterprise. It is optional: paid plans run on Since AI and its credits by default. To use your own AI provider, connect Anthropic or OpenAI on Pro. Studio and Enterprise also support AWS Bedrock and any OpenAI-compatible endpoint, such as OpenRouter or Moonshot, by its HTTPS base URL. An owner or admin turns it on under Connections, AI provider, then types any model ID your provider account can use, or picks a suggestion; the connection test checks it before saving, and Grant needs strict tool calling. One provider and model serves Grant, his team and repository AI. Your provider bills paid Grant and repository model usage directly, and that work uses no credits. If the selected provider fails, work waits for recovery instead of silently switching to Since AI. Free uses Since AI only. World State watches run on Since AI within their own monthly limits. A watched subject counts once across its aspects and attached agents.
What uses credits?
Credits measure the work Since AI does for you: Grant's answers, every teammate's check-ins and consults, scheduled reports and repository repairs. Bigger jobs use more. Free includes 1,000 credits a month, Pro 25,000 and Studio 80,000, and Enterprise the amount in its contract. They reset on the 1st (UTC) and do not roll over. When Free credits run out, Since AI pauses until the 1st or an upgrade. Pro and Studio keep going: credits past the included amount are billed automatically as extra credits at the per-1,000 rate on each plan card, on a separate monthly invoice, with no cap and nothing to switch on. If a payment is late, Pro and Studio stop at their included credits until it clears. A model call that fails uses no credits, and neither do World State watches or work on your own AI provider. Usage and billing shows credits left and the reset date, a banner appears at 80% and 100% of the included credits, and the owner gets an email when they run out.
More in the docs, or email hello@since.dev. For CI, AWS costs, security and the COO role at a small startup, read DevOps for startups and the other operations guides.
Clock out. Grant’s team is on.
The team runs check-ins on your plan’s schedule and reports back when it’s done.
The night shiftIllustrative
- Mayacheck-in (workflow run)
- Priyanew high finding · notice posted to #ops
- Lizscheduled check-in
- Noahscheduled check-in
Scheduled check-ins run daily on Pro and every six hours on Studio.