AI Driven SDLC
Login

AI Driven SDLC Documentation

Guides, references, and best practices for the AI Driven SDLC platform.

Help

Frequently Asked Questions

Short answers when something’s unclear.


Getting started

How do I create a workspace?
First login asks you to create one or join an existing org. Name it something recognizable (usually your company name). If your team already has a workspace, get an invite from an admin instead.

Can multiple people share a workspace?
Yes: that’s the point. Admins invite others from Settings → Workspace → Members. Invitees get an email with a link to accept or decline; they join the workspace only after they accept. Everyone who has joined sees the same integrations, scans, and metrics.

I got a workspace invite email — what do I do?
Open the link, sign in or create an account with the invited email, then accept (or decline). Signing up alone without accepting does not add you to the workspace.

I used email signup but want Google now. Do I lose data?
No, as long as the email matches. The accounts merge automatically.


Onboarding

Do the two onboarding steps have to be done in order?
Yes the first time: save company profile, then integrations. You can edit completed steps later from the sidebar.

I can’t install the GitHub App myself.
Ask a GitHub org owner to install it for you (Integrations setup guide). You’ll see “connected” once they’re done.

Is the HR org chart CSV required?
No. It helps team filters in Insights but onboarding completes without it.


Agentic Foundation

How long does a scan take?
Usually 5–15 minutes for normal repos. Huge monorepos can take longer; the dashboard shows progress.

How often do repos re-scan?
About every 24 hours automatically. Manual re-scan anytime from Agentic Foundation.

Why is confidence low on a dimension?
We didn’t find much evidence either way: e.g. no test files means shaky test-coverage signal. Adding the missing pieces raises confidence and often the score.

A criterion says N/A: is that bad?
No. It doesn’t apply to this repo (a docs-only repo doesn’t need a Dockerfile). N/A items don’t count against you.

What does Blocked mean?
We couldn’t verify that check at scan time, often because evidence lives outside the repo or a GitHub API call failed. It usually counts as not met until you re-scan with reachable evidence. It’s not the same as N/A. When split infrastructure is on and a selected platform repo cannot be read, those checks are left out of the percent instead of scored as a gap. See Linked platform repositories.

Insights shows failing builds, but Agentic Foundation still says CI is set up. Which is right?
Both. Agentic Foundation setup rows ask whether CI is configured (in the repo and/or linked CI templates). SDLC Insights asks whether runs have been healthy lately. Failing Actions do not rewrite CI on change, CI runs build, CI runs tests, or CI static analysis. Enable GitHub Actions if the Insights charts have no live runs. See Four CI questions.

My score dropped after a re-scan.
New gaps showed up: deleted files, rubric updates, or something that wasn’t visible last time. Check the AI Resolution Hub for fresh items.

I changed Readiness series on the overview chart. Why am I still on the same page?
That control only switches the line on the workspace chart. Default is Overall. It does not open a dimension-detail page. Use the sidebar (or a ?dim= link) for that page. The detail view has no workspace series selector. See Workspace score over time.

Activity Log is empty even though we scanned repos.
The log only lists Fix requested and PR merged events. Scan progress, scores, and run history stay on the dashboard and repo detail. Empty means nobody has requested a fix (or a fix PR hasn’t merged) yet. See Activity Log.


Insights

Why can Agentic readiness look low next to a high engineering delivery score?
They measure different things. Agentic readiness measures repo-agent maturity from scans; the engineering delivery score measures live delivery signals from connected integrations.

A workspace can show 24% Not ready next to 83/100 plus Delivery risk detected in live metrics. The /100 number is the average of live metric bands. The headline is the worst live signal, not a second overall score. The Journey hero does not invent an SDLC number, and SDLC Insights does not reuse the Agentic readiness % when there is no live delivery data. See Your AI Journey and SDLC Insights.

What is Ship Weight?
A 0 to 10 score per merged GitHub PR that estimates review load on the diff (file type, density, add vs edit vs delete), shown as a sum next to Merged PRs. 0 means the change doesn’t contribute to ship weight (docs, config, assets, or a revert). That is not the same as Backfilling. It is not raw line count, and it does not replace the PR count. See Ship Weight for what the bands mean and how it’s calculated.

Why does a PR show 0?
0 means the change doesn’t contribute to ship weight (docs, config, assets, or a revert). Hover the badge in the day drawer for the same explanation. This is different from Backfilling, which means the score isn’t ready yet.

Why does Ship Weight say Backfilling?
Scores are computed after merge. Until ingest finishes, KPIs and charts show Backfilling. Merged PR counts are unchanged; switch back to that tab if you need volume right now.

How far back does data go?
Most views default to eight weeks; widen the range filter to see older retained history. The platform keeps Insights metrics, readiness scorecards, and HR absence history for about 12 months, and AI usage daily series for about 180 days. Soft-deleted workspaces can be restored for 30 days. Full detail: Data retention.

Why “No data” or placeholder charts?
Usually: integration not connected, connected less than ~24 hours ago, or nothing happened in that window for the metric you’re viewing. Empty Dependabot or secret scanning charts also need optional GitHub Insights setup on each organization.

Why does Insights & Recommendations say Computing or Unavailable?
Computing means that window is still loading, including after you change filters. Cards show a skeleton, not Healthy or $0.00. Unavailable plus — means the section has no usable source (no CI, no merged PRs, no mapped engineers) or the load failed. After a successful load, $0.00 can be a real unit cost when merged PRs exist. See Insights & Recommendations.

Can I export CSV?
Not yet: dashboard only for now.


Integrations & privacy

Do you store our source code?
No long-term copy. We tarball a snapshot for scoring and throw it away when the scan finishes.

What does the GitHub App ask for?
AI Driven SDLC Insights (connect): read-only repo contents, metadata, PR and review events. Optional security and financial Insights metrics need extra App permissions and, for alert metrics, Code security products enabled on GitHub. See GitHub integration. AI Driven SDLC Automations (optional, automated fixes): Contents and Pull requests write so the platform can open fix PRs you review. Nothing is merged automatically.

Dependabot or secret scanning metrics are empty after connect.
The App can only read alerts. In Connect GitHub, the yellow Optional Insights metrics setup panel names which accounts still need products. Expand that account under Granted repositories and use Setup (Open Configurations on an org, Open Code security on a personal account). A GitHub org owner must enable Dependabot alerts and, if you want that metric, Secret scanning. Details: GitHub integration.

How do I disconnect something?
Settings → Integrations → Remove. Old data stays; new updates stop.


Account

Security basics?
HTTPS everywhere, encrypted integration tokens, Supabase Auth for login sessions.

Delete my account?
Open Settings → Data → Delete my account, type your email to confirm, and choose Permanently delete account. This is immediate and irreversible: it removes your account, every workspace you own, and their readiness reports, metrics history, team configuration, and integrations. Workspaces where you are only a member stay put, and your membership row is removed — analytics that other workspace admins own may still reference your work until they use Settings → Data → Erase a person’s insights data in that workspace.

Workspace admins looking to delete a whole workspace should use the workspace profile page instead of the account delete flow.

Still stuck?
Support contact is in your welcome email or admin console.