catalog roles mislabels the personality meld as "boundary:" #322

New issue

Closed

opened 2026-08-23 22:28:30 +00:00 by coilyco-ops · 1 comment

coilyco-ops commented

2026-08-23 22:28:30 +00:00

Member

Observed

agent-compose catalog roles (v2.39.0) prints the role's personality meld under the label boundary:.

platform // role-platform // Build and land the foundational software the rest of the estate is built on. // boundary: tenacious, grounded // color: #9c8b31
devrel // role-devrel // Turn real portfolio work and audience evidence into accurate content, respectful conversations, and informed commitments. // boundary: warm, outward // color: #f7ab5d

tenacious, grounded, warm, and outward are personality slugs. They are not boundaries.

Expected

The real boundaries for a role are the four boundary-* slugs the identity card carries, for example boundary-suggest-external-comms, boundary-modify-live-backend, boundary-seek-external-validation, boundary-build-foundational-software.

The field should read meld: or personality:, whichever matches the vocabulary the rest of the CLI uses.

Why it matters

The two concepts do different work. A personality tunes attention and voice, a boundary governs who is allowed to take an action. Printing one under the other's name teaches the wrong model to every reader of this command, and it is the first surface a newcomer hits.

Found while scripting a terminal demo that runs catalog roles on camera.

Scope

One label string. No data change.

## Observed `agent-compose catalog roles` (v2.39.0) prints the role's personality meld under the label `boundary:`. ``` platform // role-platform // Build and land the foundational software the rest of the estate is built on. // boundary: tenacious, grounded // color: #9c8b31 devrel // role-devrel // Turn real portfolio work and audience evidence into accurate content, respectful conversations, and informed commitments. // boundary: warm, outward // color: #f7ab5d ``` `tenacious`, `grounded`, `warm`, and `outward` are personality slugs. They are not boundaries. ## Expected The real boundaries for a role are the four `boundary-*` slugs the identity card carries, for example `boundary-suggest-external-comms`, `boundary-modify-live-backend`, `boundary-seek-external-validation`, `boundary-build-foundational-software`. The field should read `meld:` or `personality:`, whichever matches the vocabulary the rest of the CLI uses. ## Why it matters The two concepts do different work. A personality tunes attention and voice, a boundary governs who is allowed to take an action. Printing one under the other's name teaches the wrong model to every reader of this command, and it is the first surface a newcomer hits. Found while scripting a terminal demo that runs `catalog roles` on camera. ## Scope One label string. No data change.

coilyco-ops referenced this issue

2026-08-23 22:29:09 +00:00

Terminal demo: three roles, one artifact, one prompt #324

coilyco-ops referenced this issue from a commit

2026-08-23 23:06:00 +00:00

fix(cli): name the personality meld in catalog roles

coilyco-ops closed this issue

2026-08-23 23:06:00 +00:00

coilyco-ops commented

2026-08-23 23:06:14 +00:00

Author

Member

Fixed on main in 89b81ae.

The label now reads personalities:, picked over meld: because it matches the personalities JSON field on the role item and the catalog personalities subcommand, which is the vocabulary the rest of the CLI already uses.

platform // role-platform // Build and land the foundational software. // personalities: tenacious, grounded // color: #9c8b31

The format string moved into catalogRoleLine so it could be asserted, and TestCatalogRoleLineLabelsThePersonalityMeld now fails if the word boundary returns to that line. No data change, and docs/skill-catalogues.md never quoted the text form so it needed no edit.

just test green.

Fixed on `main` in `89b81ae`. The label now reads `personalities:`, picked over `meld:` because it matches the `personalities` JSON field on the role item and the `catalog personalities` subcommand, which is the vocabulary the rest of the CLI already uses. ``` platform // role-platform // Build and land the foundational software. // personalities: tenacious, grounded // color: #9c8b31 ``` The format string moved into `catalogRoleLine` so it could be asserted, and `TestCatalogRoleLineLabelsThePersonalityMeld` now fails if the word `boundary` returns to that line. No data change, and `docs/skill-catalogues.md` never quoted the text form so it needed no edit. `just test` green.

coilyco-ops referenced this issue

2026-08-24 00:23:37 +00:00

Terminal demo: three roles, one artifact, one prompt #324

coilyco-ops referenced this issue

2026-08-24 01:06:27 +00:00

No compact surface shows boundary assignment across roles #325