feat(orchestrator): markdown-backed agent loader + full integration flow#619
Open
gewenyu99 wants to merge 1 commit into
Open
Conversation
🧙 Wizard CIRun the Wizard CI and test your changes against wizard-workbench example apps by replying with a GitHub comment using one of the following commands: Test all apps:
Test all apps in a directory:
Test an individual app:
Show more apps
Results will be posted here when complete. |
This was referenced Jun 9, 2026
Collaborator
Author
|
Warning This pull request is not mergeable via GitHub because a downstack PR is open. Once all requirements are satisfied, merge this PR as a stack on Graphite.
This stack of pull requests is managed by Graphite. Learn more about stacking. |
This was referenced Jun 9, 2026
gewenyu99
force-pushed
the
experiment/orchestrator-07-walking-skeleton
branch
from
b2da91d to
f181850
Compare
gewenyu99
force-pushed
the
experiment/orchestrator-06-agent-loader
branch
from
8ff15e5 to
08821b7
Compare
gewenyu99
force-pushed
the
experiment/orchestrator-07-walking-skeleton
branch
from
f181850 to
e3fbb21
Compare
gewenyu99
force-pushed
the
experiment/orchestrator-06-agent-loader
branch
from
08821b7 to
93d7294
Compare
gewenyu99
force-pushed
the
experiment/orchestrator-07-walking-skeleton
branch
from
e3fbb21 to
fea70a9
Compare
gewenyu99
force-pushed
the
experiment/orchestrator-06-agent-loader
branch
2 times, most recently
from
4d5cb4d to
2d53010
Compare
gewenyu99
changed the base branch from
experiment/orchestrator-07-walking-skeleton
to
experiment/orchestrator-05-executor-framework
gewenyu99
force-pushed
the
experiment/orchestrator-06-agent-loader
branch
from
2d53010 to
d4d2633
Compare
gewenyu99
force-pushed
the
experiment/orchestrator-05-executor-framework
branch
from
bd9b756 to
fa90573
Compare
gewenyu99
force-pushed
the
experiment/orchestrator-06-agent-loader
branch
from
d4d2633 to
1fb8a70
Compare
The runner: a seed agent plans the task graph and an executor drains it, one fresh agent per task, independent tasks in parallel, the TUI rendering the queue — terminal skipped tasks crossed out, the agent's own task tools gated off so the queue is the sole progress surface. Both the WHAT (agent prompts: model, goal, success criteria, tools) and the HOW (mini-skills) are markdown served from context-mill, fetched once at startup into a per-flow registry: agents declare flow and seed in frontmatter, so audit and migration flows can ship their own sets without touching the wizard. Each task agent gets its own wizard-tools binding so complete_task attribution holds under parallelism. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
gewenyu99
force-pushed
the
experiment/orchestrator-06-agent-loader
branch
from
1fb8a70 to
2c61397
Compare
This was referenced Jun 10, 2026
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
The full integration flow. A seed agent plans the task graph and the executor drains it, one fresh agent per task with its own model, tools, and wizard-tools binding. Agent prompts (the WHAT) and mini-skills (the HOW) are markdown served from context-mill, fetched once at startup into a per-flow registry. Agents declare
flowandseedin frontmatter, so audit and migration ship their own sets without touching the wizard.Task prompts point at the framework's
EXAMPLE.mdandCOMMANDMENTS.mdfrom the installed reference skill. The queue panel is the sole progress surface, in the TUI and in CI logging, and skipped tasks render crossed out.