Most research is done in private and stays there. Public-sector research is the opposite. Under freedom-of-information rules, the analysis behind a decision can be requested, released, and read by a journalist, an opposition staffer, or a litigant, often long after the work is done and the team has moved on.
That changes the standard. It is not enough for the conclusion to be right. The whole process has to be disclosable: which sources were used, which were ignored, when the research was performed, and who or what produced each conclusion. If you cannot show that cleanly, an access request becomes a liability instead of a routine disclosure.
Why "we'll reconstruct it later" fails
The usual approach is to do the research, write the conclusion, and assume the trail can be reconstructed if anyone asks. In practice that reconstruction is painful and incomplete. Emails get lost. The analyst who pulled the numbers has left. The version that was actually used cannot be told apart from the three drafts around it. And generic AI tools make it worse: they cannot tell you what they looked at, what they skipped, or prove the output was not altered after the fact.
So when the request arrives, the team scrambles to assemble a chain of custody that was never captured in the first place.
Disclosable by construction
Bricolage records the trail as the research happens, not after. Every claim traces to its source. The system logs what it examined, the path it took, and when, and flags source independence and potential conflicts of interest along the way. Each run is sealed with a cryptographically signed audit receipt, so the record is tamper-evident: alter a single entry and the signature no longer verifies.
When an access request lands, the answer is already prepared. You can show the sources used and the ones set aside, the time the research ran, the methodology, and proof the record is intact, verifiable independently against a public key. The disclosure stops being a fire drill and becomes a file you already have.
Transparency as an asset, not a threat
For a public body, this flips freedom-of-information from a risk to manage into a standard you can meet by default. The research is built to be seen. A journalist could file for it and find a clean, sourced, methodology-documented record, which is exactly what a public institution should be able to produce.
The work that survives scrutiny is not the work that was done carefully and then documented from memory. It is the work that captured its own trail while it was being done.