fix(helm): accept "multibot-mentions" in allowUserMessages validation

[brettchien]

What problem does this solve?

The openab binary added MultibotMentions to the AllowUsers enum in #464, but the Helm chart's validation block in charts/openab/templates/configmap.yaml was not updated. Any user who follows docs/discord.md § Multi-Bot Setup and sets allowUserMessages: "multibot-mentions" is blocked by a template fail before the pod ever starts:

Error: UPGRADE FAILED: execution error at (openab/templates/configmap.yaml:48:8):
agents.kiro.discord.allowUserMessages must be one of: involved, mentions — got: multibot-mentions

Fixes #471
Refs #464

Discord Discussion URL: https://discord.com/channels/1491295327620169908/1491365150664560881/1495429672920416357

At a Glance

user values.yaml                         chart template (configmap.yaml)
allowUserMessages: "multibot-mentions"   has $val (list "involved" "mentions")
          │                                       │
          ▼                                       ▼
   ┌────────────┐                         ┌─────────────────────────────┐
   │ helm       │ ──── template fail ───▶ │ "must be one of: involved,  │
   │ upgrade    │                         │  mentions — got: ...-mentions"│
   └────────────┘                         └─────────────────────────────┘
                                                  ❌ blocks valid config
Binary side (same config):
   ┌────────────┐   ┌──────────────────────┐   ┌──────────────────────┐
   │ config.rs  │──▶│ AllowUsers::          │──▶│ works correctly      │
   │ deserialize│   │ MultibotMentions     │   │ (#464 discord.rs /   │
   └────────────┘   └──────────────────────┘   │  slack.rs branches)  │
                                                └──────────────────────┘

Prior Art & Industry Research

• OpenClaw — ACP bind layer, no Helm chart with allow-mode validation. No analogue.
• Hermes Agent — skills-based, no persistent service config validation. No analogue.
• openab itself — allowBotMessages already validates with has ... (list "off" "mentions" "all") and has corresponding unit tests. This PR aligns allowUserMessages with that existing pattern.

Proposed Solution

Add "multibot-mentions" to the has allow-list for both Discord and Slack, plus update the documentation comment and charts/openab/values.yaml user-facing docs. Mirror the existing allowBotMessages test pattern for allowUserMessages.

-    {{- if not (has $cfg.discord.allowUserMessages (list "involved" "mentions")) }}
-    {{- fail (printf "agents.%s.discord.allowUserMessages must be one of: involved, mentions — got: %s" $name $cfg.discord.allowUserMessages) }}
+    {{- if not (has $cfg.discord.allowUserMessages (list "involved" "mentions" "multibot-mentions")) }}
+    {{- fail (printf "agents.%s.discord.allowUserMessages must be one of: involved, mentions, multibot-mentions — got: %s" $name $cfg.discord.allowUserMessages) }}

Same one-line change at the Slack block. Added 4 new helm-unittest cases covering involved / mentions / multibot-mentions renders plus the invalid-value rejection (the test file previously only covered allowBotMessages — allowUserMessages had no test coverage).

Why this approach

• Symmetric with the existing allowBotMessages pattern — same validation style, same test style, no architectural change.
• Smallest surface that fixes the bug — no refactor, no new abstraction, pure catch-up.
• Keeps the helpful error message — users who typo the value still get a useful must be one of: ... message, now accurate.

Alternatives Considered

Alternative	Rejected because
Drop the chart validation entirely and rely on binary-side deserializer errors	Loses the template-time feedback (fails after pod start rather than before); inconsistent with existing allowBotMessages check.
Use a regex instead of has ... list	Less readable; the existing pattern is a list; matching existing style is simpler.
Only accept hyphen form ("multibot-mentions") and leave underscore ("multibot_mentions") as binary-only	Chart is the user-facing surface; accepting only hyphen form matches docs/discord.md and keeps errors clear. Rust deserializer still accepts both internally.

Validation

• helm template openab ./charts/openab --set agents.kiro.discord.allowUserMessages=multibot-mentions renders allow_user_messages = "multibot-mentions" successfully.
• helm template ... --set allowUserMessages=bogus still fails with the updated error message listing all three valid values.
• helm template ... --set allowUserMessages=mentions still renders the pre-existing behaviour (backward-compatible).
• helm unittest charts/openab → 18 passed, 0 failed (including 4 new test cases for allowUserMessages).

$ helm unittest charts/openab

### Chart [ openab ] charts/openab

 PASS  adapter enablement gating	charts/openab/tests/adapter-enablement_test.yaml
 PASS  allowBotMessages & trustedBotIds	charts/openab/tests/configmap_test.yaml

Charts:      1 passed, 1 total
Test Suites: 2 passed, 2 total
Tests:       18 passed, 18 total

[thepagent]

taking a look

[chaodu-agent]

LGTM ✅

Reviewed the full diff across 3 commits:

• Core fix: multibot-mentions added to allow-list for both Discord and Slack validation
• Error messages now hint about hyphen form to prevent underscore copy-paste footgun
• Test coverage: 5 new cases (Discord × 4 + Slack × 1), all 19 pass
• values.yaml docs updated with clear mode descriptions

CI green, no nits remaining. Ship it 🚀