assessing-an-anchor
Use when evaluating a potential anchor or BD pipeline partner that isn't integrated yet — researching it against the scorecard signals and adding it as an in-vetting honorable mention, without writing
SKILL.md.claude/skills/assessing-an-anchor/SKILL.md
GitHubAssessing a potential anchor (vetting)
For a candidate not yet integrated — typically a BD pipeline partner — assess it without writing client code.
- Research against the 9 scorecard criteria (
QUALITY_CRITERIAinsrc/lib/config/anchors.ts). Establish the gating fact first: does it support Stellar at all (anchor / SEP / USDC-on-Stellar / via a partner)? Then score each signalmet/partial/failed/unverifiedwith evidence. A USD-denominated token (USDC) does not satisfylocal-asset— only a local-currency token on Stellar does. Parallel research subagents (one per anchor) work well. Demand sources, and preferunverifiedover guessing. - Add it as an in-vetting honorable mention in
HONORABLE_MENTIONSwithvetting: trueand ascorecardbuilt viamakeCriteria()from the research. It then flows automatically into/api/scorecardand the scorecard page, with the "Under evaluation" badge, for BD tracking. - New markets not in
regions.tsare fine: they render in the scorecard only, with no region page, following the Philippines precedent. But any region or rail id missing fromregions.ts/rails.tsmust be baselined inKNOWN_DANGLINGintests/config/config-integrity.test.ts, or that test fails. - Cover with config tests —
tests/config/anchors.test.tsfor counts and membership,tests/config/scorecard.test.tsfor verdict and vetting. Update the entry's scorecard as the partner makes progress.
Promotion out of vetting is an editorial call, not a computed one; see .claude/rules/curation-and-scorecard.md. When a candidate graduates, follow the adding-an-anchor skill.