feat(omp): add native oh-my-pi harness support#1996
Open
lucasmedeirosleite wants to merge 1 commit into
Open
Conversation
lucasmedeirosleite
force-pushed
the
feat/add-omp-support
branch
from
July 23, 2026 18:08
22c73d7 to
b68ac9a
Compare
lucasmedeirosleite
force-pushed
the
feat/add-omp-support
branch
from
July 24, 2026 16:28
b68ac9a to
9cba706
Compare
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Who is submitting this PR? (required)
Installed plugins:
What problem are you trying to solve?
Oh-my-pi users currently do not get a first-class Superpowers integration. The plugin can ship skills, but a clean omp session does not get the
using-superpowersbootstrap injected at sessionstart, so the skills are present without the behavior that makes them auto-trigger.
This addresses #1941.
What does this PR change?
Adds a native oh-my-pi adapter through
package.json#omp.extensions, backed by a shared harness-neutral bootstrap controller used by both pi and omp. Pi keeps its existing liveresources_discoverbehavior; OMP uses its native package-rootskills/discovery and registers only the lifecycle hooks omp actually emits.Also adds omp install/verification docs, harness-porting guidance, and contract tests for the shared controller, pi adapter, omp adapter, and omp docs.
Is this change appropriate for the core library?
Yes. This is new harness support for a general-purpose agent runner, which is the same kind of core integration this repo already ships for other harnesses. It does not add a domain-specific
skill, project-specific configuration, or a third-party runtime dependency.
The shared controller is deliberately narrow: it owns only harness-neutral bootstrap policy — arm/disarm, message placement, deduplication, frontmatter stripping, caching, and fail-soft
diagnostics. The Pi and OMP adapters still own their own manifests, lifecycle hooks, discovery boundaries, logging, and tool mappings.
What alternatives did you consider?
skill://skill reads, lowercasetask, and lowercasetodo.resources_discoverevent. and rejected after checking omp 16.5.2 source. The type exists, but there are no runtime emit callsites. omp's actual native skill discovery forextension packages scans the plugin root's conventional
skills/directory, so the omp adapter does not register a dormant hook.omp.extensions; the docs use omp's Git installer and localomp plugin linkpathinstead.
Does this PR contain multiple unrelated changes?
No. The changes are all part of one harness-support slice: shared bootstrap policy, the native omp adapter, the pi refactor needed to share that policy safely, docs, and tests.
Existing PRs
Searches used:
omp,oh-my-pi,oh my pi,native omp, and1941across open and closed PRs.Environment tested
gpt-5.6-solgpt-5.6-solNew harness support (required if this PR adds a new harness)
Clean-session transcript for "Let's make a react todo list"
Evaluation
fallback.
adapter, and OMP docs.
a clarifying question.
Verification run:
Results:
Rigor
Adversarial checks included:
Human review