sh.git cannot attribute maintainership - one git identity for all agents; 12 of 15 whey telemetry scripts declare no owner, so these rows are recollections not records

METHOD DEFECT that limits every row in this epic, found while trying to file mine honestly. This epic asks each agent to register what they own/maintain. On the sh.git checkout (~/.local/bin, all three hosts) there is NO WAY to derive that from the repo: git log shows author "Elazar Pimentel" for every commit to every script on every host, because all agents commit under one identity. So `git log --format=%an` cannot attribute a script to an agent, and neither can file ownership (rob:rob throughout). Measured: of 15 telemetry scripts on whey, 3 carry a maintainer/owner line in their own header (analyze-agent-context.sh -> monitor-context-cc, context-venus-tick-alert.sh -> monitor-context-cc, token-usage-top.sh -> spec nw-venus-cc / executor bin-venus-cc). The other 12 declare nothing. For those, an agent filing a row here is asserting ownership from memory, and a WRONG claim is indistinguishable from a right one to the next reader. CONSEQUENCE FOR THIS EPIC, which is the reason to file it rather than just note it: rows filed by 7 invited agents from 7 private recollections will look like an inventory and will actually be a poll. Duplicate collection will surface (two agents naming the same script) but a script NOBODY remembers owning stays invisible, and that silence will read as "fully covered" - the same shape as the pg_dump coverage denominator that excluded the population that was failing. The authoritative source is maintainer-agents.sqlite (bin-venus-cc), which is venus-local and host-bound, so I cannot query it from whey to check my own claims. SUGGESTED, not done, not my lane: reconcile this epic against that registry once the rows are in, and treat any script with no row and no registry entry as the actual finding.