The flagship deliverable
One evidence-based picture for the people who decide.
The Diagnostic Package is a structured account of what was observed, what deserves investigation, how confident we are, and what remains unknown, assembled for an executive and an architect to read together and act on. It is not a report, a dashboard, a scorecard, or a verdict. It is decision intelligence.
Why it exists
Operational tooling answers operational questions. The decision needs something else.
Most tooling was built to answer operational questions, is it running, is it secure, is the build green. None of it was built for the moment of decision, when a leadership team must commit to changing software it does not fully understand.
Dashboards report status; scanners report issues. Neither tells you where understanding is thin, where investigation should begin, or how confident anyone should be. They describe the system, they do not inform the decision.
A decision is made by people with different vantage points. The package gives an executive and an architect the same evidence, the same priorities, and the same honest account of the unknowns, so they decide from one picture, not three opinions.
The Diagnostic Package exists for the decision, and is honest about the limits of what it knows.
What’s inside
Each component shows what RepoTruth sees, what it interprets, and what only your people can decide.
A leadership-readable account, what was observed, what deserves investigation, and what remains unknown, in one sitting.
Surfaces an evidence-based orientation for decision-makers.
Does not approve, reject, or recommend a course of action.
The system’s actual structure and strata, modules, boundaries, dependency concentration, and how it formed over time, observed rather than assumed.
Surfaces the real shape of the system, mapped from evidence.
Does not grade the architecture or declare it good or bad.
Every observation, cited to its source, each carrying its own confidence, the auditable base everything else rests on.
Surfaces traceable evidence a reviewer can check.
Does not present inference as fact.
Where to investigate first, ordered by investigation urgency and paired with confidence and unknown density.
Surfaces investigation urgency, where human attention is best spent first.
Does not recommend actions, or rank by severity.
What could not be determined, ownership, runtime behavior, rationale, surfaced explicitly and with equal prominence.
Surfaces the questions worth asking next.
Does not hide gaps or infer them away.
How firmly each finding is held, and why, decomposed into its evidence, never a single opaque number.
Surfaces explainable confidence for every observation.
Does not produce black-box scores.
The evidence made visible, dependency concentration, knowledge thinning, the fog of unknowns, drillable to the evidence beneath.
Surfaces a legible visual reading of the evidence.
Does not issue a score, grade, or verdict.
A plain statement of exactly what was analyzed, what was not, what remains unknown, and the boundaries RepoTruth operates within.
Surfaces the package’s own limits, in writing.
Does not overclaim, certify, or imply completeness.
Inside the Executive Report
Built so the people who decide and the people who design can read from the same evidence.
The shape of the decision: what was observed, what deserves the earliest investigation, and what is still unknown, in language a leadership team can act on.
Structural reality and its concerns, chokepoints, dependency concentration, boundaries, each as a candidate for human judgment, never a verdict.
An explicit account of the questions the snapshot could not answer, carried into the report rather than left out of it.
The Unknown Inventory
An unknown isn’t a gap to hide. It’s one of the most useful things we can give you.
Most tools quietly fill in what they cannot determine, or omit it entirely. RepoTruth records it. Where the evidence runs out, the package names an Unknown, explicitly, prominently, and as a first-class finding.
An unknown is not low risk. It is un-assessed. A region full of unknowns is not safe, it is precisely where your people should look. Often the most valuable page in the package is the one that lists, exactly, the questions worth asking next.
We would rather tell you the truth about what we can’t see than hand you a false sense of completeness.
Confidence Matrix
No black-box conclusions. Nothing asserted more firmly than the evidence allows.
What was observed, cited to its source, the basis a reviewer can independently check.
How firmly it is held, on an explainable scale, decomposed into its reasons, never a single opaque number.
What would change the finding if resolved, the decay and the unknowns stated alongside it.
Confidence describes how firmly we know something. It never describes whether you should act.
Constitutional Manifest
The most honest page is the one that draws the boundary.
Every Diagnostic Package ships with a Constitutional Manifest, a plain account a security or legal team can read, declaring exactly where the analysis began and ended.
A package you can trust is one that tells you where its knowledge ends.
Evidence Visualization Layer
The Observatory Projection makes the package’s evidence visible, the visualization layer beneath the findings, not a verdict.
It reveals where dependency concentrates, where change churns, where ownership is uncertain, where knowledge thins, and where the unknowns cluster, the regions that most deserve a human’s closest reading, and stays drillable to the evidence beneath.
It shows; it assigns no grades, computes no score, and decides nothing. Your people interpret it.
In practice
One deliverable, used before the moments that are hardest to reverse.
Understand where automated change carries the most risk, so human review precedes it.
Identify where change is likely to propagate, and where to look before you move.
See the real structure, and its unknowns, before committing a multi-year program.
Establish the observed baseline before you redraw it.
Build a defensible, read-only evidence base for the diligence file.
Form an independent technical view of the target, without touching the asset.
Assess delivered software objectively, on evidence rather than assertion.
Constitutional guarantees
Six permanent guarantees, properties of how RepoTruth is built, not settings you take on faith.
We write to nothing in your environment. There is no path by which RepoTruth can change your code.
We analyze an inert, point-in-time copy, never a live or running system.
We never connect to production. Not to observe, not for any reason.
We store, request, and reuse no credentials, because we connect to nothing that needs them.
We execute, deploy, and remediate nothing. RepoTruth cannot act on your systems.
Every output is built for a human to review and decide on. RepoTruth reaches no conclusion of its own.
See it for yourself
Before you share anything of your own, you can walk through a complete Diagnostic Package assembled on a synthetic codebase, executive report, atlas, evidence catalog, review priorities, unknown inventory, confidence matrix, and constitutional manifest. It is entirely synthetic: no customer artifacts, no real repository, no invented metrics.
Request the sample packageProvided on request. Synthetic demonstration only.
A read-only diagnostic, scoped to your decision.
Tell us the decision you’re facing. We’ll scope a Diagnostic Package to it, read-only, snapshot-based, with no production access and no obligation to act on what we find. Your people will know where to look first, how confident we are, and what is still unknown.
No credentials. No production access. We never change your code.
