Skip to content

docs(maxsim-batch): add tier selection table and graceful degradation (§7.2) #188

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
maystudios merged 1 commit into from
Mar 25, 2026
Merged

docs(maxsim-batch): add tier selection table and graceful degradation (§7.2) #188

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
22 changes: 22 additions & 0 deletions templates/skills/maxsim-batch/SKILL.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -89,12 +89,34 @@ Agent Teams (available since Claude Code v2.1.32, Feb 2026) enable inter-agent c

**Current status:** Infrastructure is in place (env var, hooks). Workflow templates that invoke `TeamCreate`/`SendMessage` for Tier 2 patterns (competitive implementation, multi-reviewer code review, collaborative debugging) are planned but not yet implemented. All workflows currently use Tier 1 subagents. See PROJECT.md §7.2 for the full specification.

### Tier Selection Logic

MaxsimCLI chooses the tier automatically based on the workflow:

Comment on lines +92 to +95
Copy link

Copilot AI Mar 25, 2026

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

The new Tier Selection Logic section states that “MaxsimCLI chooses the tier automatically based on the workflow”, but this conflicts with the preceding “Current status” note that Tier 2 workflow templates are not yet implemented and all workflows currently run Tier 1. Consider clarifying that this table describes the intended/spec selection logic once Tier 2 templates exist (or otherwise reconcile these statements).

Copilot uses AI. Check for mistakes.
| Workflow | Tier | Reason |
|----------|------|--------|
| Phase execution (independent tasks) | Tier 1 (Subagents) | Tasks don't need to communicate |
| Codebase scanning | Tier 1 (Subagents) | Read-only, report back |
| Research gathering | Tier 1 (Subagents) | Collect and report |
| Competitive implementation | Tier 2 (Agent Teams) | Agents need to debate |
| Multi-dimensional code review | Tier 2 (Agent Teams) | Findings need cross-checking |
| Collaborative debugging | Tier 2 (Agent Teams) | Hypotheses need adversarial testing |
| Architecture exploration | Tier 2 (Agent Teams) | Requires discussion |

**When Tier 2 is ready, it will be used for:**
- Competitive implementation with adversarial debate
- Multi-dimensional code review (security + performance + test coverage)
- Collaborative debugging with competing hypotheses
- Cross-layer feature work (frontend + backend + tests)

### Graceful Degradation

If Agent Teams are unavailable (env var `CLAUDE_CODE_EXPERIMENTAL_AGENT_TEAMS` not set, unsupported plan, or feature not yet stable), MaxsimCLI falls back to Tier 1 subagents for all workflows. Inform the user with this exact message:

> "Competitive mode: using Tier 1 subagents (Agent Teams not available or not required for this strategy). Each executor works independently; verifier selects the best result."

Comment on lines +114 to +117
Copy link

Copilot AI Mar 25, 2026

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

The Graceful Degradation section says Tier 1 fallback applies “for all workflows” and instructs using an “exact message”, but the message itself is specific to “Competitive mode”. If this message is only meant for competitive implementation fallback (as in the execute workflow template), consider scoping this section/message to competitive mode, and/or documenting the user-facing message(s) for other workflow degradations separately to avoid misleading guidance.

Suggested change
If Agent Teams are unavailable (env var `CLAUDE_CODE_EXPERIMENTAL_AGENT_TEAMS` not set, unsupported plan, or feature not yet stable), MaxsimCLI falls back to Tier 1 subagents for all workflows. Inform the user with this exact message:
> "Competitive mode: using Tier 1 subagents (Agent Teams not available or not required for this strategy). Each executor works independently; verifier selects the best result."
If Agent Teams are unavailable (env var `CLAUDE_CODE_EXPERIMENTAL_AGENT_TEAMS` not set, unsupported plan, or feature not yet stable), MaxsimCLI falls back to Tier 1 subagents for all workflows.
For **competitive implementation / competitive mode** workflows, inform the user with this exact message:
> "Competitive mode: using Tier 1 subagents (Agent Teams not available or not required for this strategy). Each executor works independently; verifier selects the best result."
For other workflows, clearly inform the user that Agent Teams are unavailable and that Tier 1 subagents are being used instead, using wording appropriate to that workflow.

Copilot uses AI. Check for mistakes.
The user is informed but not blocked. All workflows remain fully functional via Tier 1.

## Limits

- Up to 30 parallel agents; typically 3-10 for manageable coordination
Expand Down
Loading