cleveragents/cleveragents-core

Fork 3

agents actor list returns empty after fresh agents init — built-in LLM actors not seeded or auto-resolved #4215

New issue

Open

opened 2026-04-07 09:13:33 +00:00 by hurui200320 · 2 comments

hurui200320 commented

2026-04-07 09:13:33 +00:00

Member

Metadata

Branch: fix/actor-list-empty-after-init
Commit Message: fix(actor): seed or auto-resolve built-in LLM actors on agents init
Milestone: None (backlog)
Parent Epic: TBD - requires manual linking

Description

Background

After a fresh agents init --yes (exits 0, success), running agents actor list
immediately returns No actors configured. No built-in LLM actors exist for
anthropic/, openai/, gemini/, or any other provider. The spec describes
built-in actors with <provider>/<model> naming as a core feature (e.g.
anthropic/claude-3.5, openai/gpt-4), but they are absent from a freshly
initialized database.

Current Behavior

agents init --yes
# exit 0 — success

agents actor list
# Output: No actors configured.

No built-in LLM actors are present. To reference any actor in a plan
(agents plan use --execution-actor <NAME> or in action YAML), users must
first manually run agents actor add --config <file> — a prerequisite that
is not documented in the agents init output or help text.

Expected Behavior

After agents init, the actor registry should either:

(a) Auto-seed a set of built-in LLM actors for each configured provider,
so agents actor list shows them immediately after init, OR
(b) Auto-resolve <provider>/<model> names at runtime without requiring
explicit actor add, and document this clearly so actor list output is not
misleading.

Currently neither (a) nor (b) is implemented, leaving users unable to run any
plan without first understanding the undocumented actor add prerequisite step.

Spec references:

Naming & Identity: "Built-in LLM actors use provider prefixes (e.g.,
openai/, anthropic/). Built-in resource types are unnamespaced."
agents actor run synopsis: agents actor run ... <NAME> <PROMPT> — where
<NAME> would be e.g. anthropic/claude-3.5
agents session create example: agents session create --actor local/orchestrator
— implies actors are referenceable by name without explicit actor add

Acceptance Criteria

After agents init --yes, agents actor list shows at least the built-in
LLM actors for all configured providers (option a), OR agents actor list
clearly indicates that provider-prefixed actors are auto-resolved at runtime
and do not require explicit registration (option b)
agents plan use --execution-actor anthropic/claude-3.5 (or equivalent)
works without a prior agents actor add step
The agents init success output or help text documents the actor
prerequisite (if actor add is still required)
No regression in existing actor add / actor list behaviour

Subtasks

Investigate whether the spec mandates option (a) seeding or option (b)
runtime auto-resolution, and confirm the chosen approach with the team
Implement the chosen approach: either seed built-in actors during
agents init or add runtime auto-resolution logic in the actor registry
Update agents init output / help text to document actor setup if
actor add remains a required step
Tests: add/update tests covering the change (unit + integration for
agents init → agents actor list flow)
Run nox (all default sessions), fix any errors
Verify coverage >= 97% via nox -s coverage_report

Definition of Done

All subtasks above are completed and checked off
A Git commit is created where the first line of the commit message
matches the Commit Message in Metadata exactly
The commit is pushed to the branch from Metadata
PR submitted to master, reviewed, and merged
All nox stages pass
Coverage >= 97%

## Metadata - **Branch**: `fix/actor-list-empty-after-init` - **Commit Message**: `fix(actor): seed or auto-resolve built-in LLM actors on agents init` - **Milestone**: None (backlog) - **Parent Epic**: TBD - requires manual linking ## Description ### Background After a fresh `agents init --yes` (exits 0, success), running `agents actor list` immediately returns `No actors configured.` No built-in LLM actors exist for `anthropic/`, `openai/`, `gemini/`, or any other provider. The spec describes built-in actors with `<provider>/<model>` naming as a core feature (e.g. `anthropic/claude-3.5`, `openai/gpt-4`), but they are absent from a freshly initialized database. ### Current Behavior ```bash agents init --yes # exit 0 — success agents actor list # Output: No actors configured. ``` No built-in LLM actors are present. To reference any actor in a plan (`agents plan use --execution-actor <NAME>` or in action YAML), users must first manually run `agents actor add --config <file>` — a prerequisite that is not documented in the `agents init` output or help text. ### Expected Behavior After `agents init`, the actor registry should either: - **(a) Auto-seed** a set of built-in LLM actors for each configured provider, so `agents actor list` shows them immediately after init, OR - **(b) Auto-resolve** `<provider>/<model>` names at runtime without requiring explicit `actor add`, and document this clearly so `actor list` output is not misleading. Currently neither (a) nor (b) is implemented, leaving users unable to run any plan without first understanding the undocumented `actor add` prerequisite step. **Spec references:** - *Naming & Identity*: "Built-in LLM actors use provider prefixes (e.g., `openai/`, `anthropic/`). Built-in resource types are unnamespaced." - `agents actor run` synopsis: `agents actor run ... <NAME> <PROMPT>` — where `<NAME>` would be e.g. `anthropic/claude-3.5` - `agents session create` example: `agents session create --actor local/orchestrator` — implies actors are referenceable by name without explicit `actor add` ### Acceptance Criteria - [ ] After `agents init --yes`, `agents actor list` shows at least the built-in LLM actors for all configured providers (option a), OR `agents actor list` clearly indicates that provider-prefixed actors are auto-resolved at runtime and do not require explicit registration (option b) - [ ] `agents plan use --execution-actor anthropic/claude-3.5` (or equivalent) works without a prior `agents actor add` step - [ ] The `agents init` success output or help text documents the actor prerequisite (if `actor add` is still required) - [ ] No regression in existing `actor add` / `actor list` behaviour ## Subtasks - [ ] Investigate whether the spec mandates option (a) seeding or option (b) runtime auto-resolution, and confirm the chosen approach with the team - [ ] Implement the chosen approach: either seed built-in actors during `agents init` or add runtime auto-resolution logic in the actor registry - [ ] Update `agents init` output / help text to document actor setup if `actor add` remains a required step - [ ] Tests: add/update tests covering the change (unit + integration for `agents init` → `agents actor list` flow) - [ ] Run `nox` (all default sessions), fix any errors - [ ] Verify coverage >= 97% via `nox -s coverage_report` ## Definition of Done - [ ] All subtasks above are completed and checked off - [ ] A Git commit is created where the first line of the commit message matches the Commit Message in Metadata exactly - [ ] The commit is pushed to the branch from Metadata - [ ] PR submitted to `master`, reviewed, and merged - [ ] All nox stages pass - [ ] Coverage >= 97%

HAL9000 commented

2026-04-08 11:17:49 +00:00

Owner

Issue reviewed and triaged.

This issue is well-formed: it has clear background, expected behavior (with two implementation options), acceptance criteria, complete metadata (commit message + branch), subtasks, and a Definition of Done.

Priority: Medium — agents actor list returning empty after init is confusing but users can work around it with actor add. However, it contradicts the spec's description of built-in actors.
Story Points: 3 (M) — requires investigating the spec intent (seed vs. auto-resolve), implementing the chosen approach, and updating documentation.
Note: This issue is related to #4300 and #4321 (actor configuration issues). The chosen approach here may influence how those fixes are implemented.
Next step: This issue is now verified and ready for implementation.

Transitioning to State/Verified.

Automated by CleverAgents Bot
Supervisor: Human Liaison | Agent: human-liaison

Issue reviewed and triaged. This issue is well-formed: it has clear background, expected behavior (with two implementation options), acceptance criteria, complete metadata (commit message + branch), subtasks, and a Definition of Done. - **Priority**: Medium — `agents actor list` returning empty after init is confusing but users can work around it with `actor add`. However, it contradicts the spec's description of built-in actors. - **Story Points**: 3 (M) — requires investigating the spec intent (seed vs. auto-resolve), implementing the chosen approach, and updating documentation. - **Note**: This issue is related to #4300 and #4321 (actor configuration issues). The chosen approach here may influence how those fixes are implemented. - **Next step**: This issue is now verified and ready for implementation. Transitioning to `State/Verified`. --- **Automated by CleverAgents Bot** Supervisor: Human Liaison | Agent: human-liaison

HAL9000 commented

2026-04-08 11:32:35 +00:00

Owner

Issue triaged by project owner:

State: Verified ✅ — Clear bug: actor list returns empty after fresh init, built-in actors not seeded
Priority: Medium ✅ (already set) — Affects first-run experience but workaround exists (manually add actors)
Milestone: v3.5.0 (M6: Autonomy Hardening) — Actor system is core to autonomy
Story Points: 3 — M — Requires implementing actor seeding/auto-resolution during init
MoSCoW: Should Have — Important for usability but actors can be added manually
Assignee: HAL9000

Automated by CleverAgents Bot
Supervisor: Project Owner | Agent: project-owner

Issue triaged by project owner: - **State**: Verified ✅ — Clear bug: actor list returns empty after fresh init, built-in actors not seeded - **Priority**: Medium ✅ (already set) — Affects first-run experience but workaround exists (manually add actors) - **Milestone**: v3.5.0 (M6: Autonomy Hardening) — Actor system is core to autonomy - **Story Points**: 3 — M — Requires implementing actor seeding/auto-resolution during init - **MoSCoW**: Should Have — Important for usability but actors can be added manually - **Assignee**: HAL9000 --- **Automated by CleverAgents Bot** Supervisor: Project Owner | Agent: project-owner

HAL9000 added this to the v3.5.0 milestone

2026-04-08 11:33:18 +00:00

HAL9000 added the