Make ticket and epic execution runtime agnostic

CHANGELOG.md

@@ -4,9 +4,11 @@ summary: Chronological history of repository and skill changes.
 
 # Changelog
 
-## 2026-07-20 — Composed ticket and epic execution
+## 2026-07-20 — Portable ticket and epic execution
 
+- feat: make ticket and epic execution runtime agnostic
 - feat: compose epic execution through implement-ticket
+  (7c4e500a35d48b5dba311094b4d34d8ca97f25a1)
 
 ## 2026-07-19 — Epic workflow and review contract cleanup
 

README.md

@@ -9,7 +9,7 @@ A personal monorepo for agent skills and supporting scripts.
   evaluation fixtures shared by repository-owned review skills
 - `justfile` — common tasks for testing, validation, and formatting
 
-Current skills:
+Current reusable agent skills:
 
 - `skills/implement-ticket` — implement exactly one standalone ticket or named
   epic child through isolated execution, repository-owned review, PR gates, and
@@ -29,6 +29,18 @@ Current skills:
 - `skills/review-code-simplicity` — reduce local cognitive load through
   behavior-preserving reuse, DRY, control-flow, and test simplification
 
+The composed implementation dependency chain is:
+
+```text
+implement-epic
+└── implement-ticket
+    └── review-code-change
+```
+
+Compatible runtimes may provide named subagents or equivalent isolated
+implementation and review contexts. OpenAI-facing files under `agents/` are
+optional discovery metadata, not part of the skills' portable contracts.
+
 ## Quick Start
 
 Run the core checks:
@@ -52,7 +64,7 @@ Validate a review packet and result together:
 python3 review-suite/scripts/validate.py pair packet.json result.json
 ```
 
-Run deterministic local evals (no Codex required):
+Run deterministic local evals without an agent runtime:
 
 ```bash
 just eval-prepare-changesets

skills/implement-epic/SKILL.md

@@ -21,6 +21,33 @@ inline a copy of the ticket workflow, or weaken any gate.
 cleanup, and terminal evidence. Do not invoke individual review lenses or
 `review-code-change` directly from this skill.
 
+## Require compatible runtime capabilities
+
+A compatible agentic runtime must be able to:
+
+- load `implement-epic` and repository-owned `implement-ticket` by stable skill
+  name or an equivalent repository-owned dependency mechanism;
+- read repository instructions plus structured GitHub or Linear relationships;
+- retain task state while repeating the dependency-driven graph loop;
+- invoke one mutating ticket execution in an exclusively owned branch/worktree
+  context and wait for its terminal result;
+- inspect and verify returned candidate, PR, tracker, and base evidence;
+- poll or wait for asynchronous ticket, CI, review, and graph-transition gates;
+  and
+- preserve a fresh read-only reviewer context through the ticket result without
+  taking ownership of local review.
+
+The portable dependency chain is `implement-epic` → `implement-ticket` →
+`review-code-change`. Verify `implement-ticket` directly and require its result
+to prove that its own review dependency and applicable capabilities were
+available. Do not make this skill invoke `review-code-change` itself.
+
+Stop before child mutation with an explicit limitation when an applicable
+capability or dependency is unavailable. Product-specific discovery metadata
+such as `agents/openai.yaml` may exist, but it does not constrain the operating
+contract or require a particular agent product. Terms such as worker and
+subagent describe possible isolated execution roles, not required product APIs.
+
 ## Load graph and closeout references
 
 - Read [the GitHub graph adapter](references/github.md) whenever GitHub owns

skills/implement-epic/evals/cases.json

@@ -39,6 +39,7 @@
     "id": "exclusive-delegated-worktree",
     "request": "Delegate ready child G-161 from epic G-160.",
     "graph": "G-161 is the only ready child.",
+    "runtime_profile": "Repository-owned skills and named subagents are available.",
     "delegation": "Worker returns a candidate from a named exclusive branch/worktree plus terminal evidence."
   },
   {
@@ -74,5 +75,31 @@
     "request": "Implement ready children G-221 and G-222 in parallel.",
     "graph": "Both are graph-independent, but both edit the same shared contract and have ordered rollout semantics.",
     "authority": "parallel execution requested"
+  },
+  {
+    "id": "equivalent-isolated-context-profile",
+    "request": "Implement epic G-230 through authorized child merges.",
+    "graph": "G-231 is the only open ready child.",
+    "runtime_profile": "The runtime calls isolated units jobs rather than subagents, but provides one exclusive mutating job, a fresh read-only audit job inside implement-ticket, retained loop state, and asynchronous waiting.",
+    "ticket_results": ["G-231 merged with internally consistent evidence"]
+  },
+  {
+    "id": "missing-review-dependency-through-ticket",
+    "request": "Implement epic G-240.",
+    "graph": "G-241 is the only open ready child.",
+    "runtime_profile": "implement-ticket is available, but repository-owned review-code-change cannot be loaded.",
+    "ticket_results": ["G-241 blocked before mutation because review-code-change is unavailable"]
+  },
+  {
+    "id": "missing-isolation-capability",
+    "request": "Implement epic G-250.",
+    "graph": "G-251 is the only open ready child.",
+    "runtime_profile": "The runtime cannot create or verify an exclusively owned mutating ticket context."
+  },
+  {
+    "id": "missing-asynchronous-wait",
+    "request": "Implement epic G-260 through child merges.",
+    "graph": "G-261 is open and ready, with asynchronous CI and review gates required by its ticket workflow.",
+    "runtime_profile": "The runtime cannot poll, wait, or resume after asynchronous child gates."
   }
 ]

skills/implement-epic/evals/results.json

@@ -63,5 +63,25 @@
     "case_id": "parallel-nonoverlap-required",
     "workflow_state": "serial_execution_required",
     "required_actions": ["reject parallel mutation", "sequence shared-contract work", "recompute readiness after each result"]
+  },
+  {
+    "case_id": "equivalent-isolated-context-profile",
+    "workflow_state": "epic_children_merged",
+    "required_actions": ["accept equivalent isolated context terminology", "verify G-231 merged evidence", "refresh the complete graph", "preserve review ownership inside implement-ticket"]
+  },
+  {
+    "case_id": "missing-review-dependency-through-ticket",
+    "workflow_state": "blocked",
+    "required_actions": ["preserve the G-241 blocked result", "name review-code-change as unavailable", "do not invoke review directly", "do not count G-241 complete"]
+  },
+  {
+    "case_id": "missing-isolation-capability",
+    "workflow_state": "blocked",
+    "required_actions": ["name exclusive mutating context isolation as unavailable", "perform no child mutation", "do not weaken ownership checks"]
+  },
+  {
+    "case_id": "missing-asynchronous-wait",
+    "workflow_state": "blocked",
+    "required_actions": ["name asynchronous wait as unavailable", "perform no child mutation", "do not pretend ticket or remote gates passed"]
   }
 ]

skills/implement-epic/scripts/tests/test_orchestration_contract.py

@@ -25,6 +25,10 @@ def setUpClass(cls):
         cls.linear = read(SKILL_ROOT / "references" / "linear.md")
         cls.closeout = read(SKILL_ROOT / "references" / "closeout.md")
         cls.contract = compact(cls.skill + cls.github + cls.linear + cls.closeout)
+        cls.eval_contract = compact(
+            read(SKILL_ROOT / "evals" / "cases.json")
+            + read(SKILL_ROOT / "evals" / "results.json")
+        )
         cls.cases = {
             item["id"]: item
             for item in json.loads(read(SKILL_ROOT / "evals" / "cases.json"))
@@ -43,6 +47,28 @@ def test_canonical_name_and_metadata(self):
         self.assertIn("$implement-epic", metadata)
         self.assertNotIn("$implement-epic-sequence", metadata)
 
+    def test_product_neutral_runtime_contract(self):
+        self.assertNotIn("Codex", self.contract)
+        self.assertNotIn("OpenAI", self.contract)
+        self.assertNotIn("Codex", self.eval_contract)
+        self.assertNotIn("OpenAI", self.eval_contract)
+        self.assertIn("compatible agentic runtime", self.contract)
+        self.assertIn(
+            "`implement-epic` and repository-owned `implement-ticket` by stable skill name",
+            self.contract,
+        )
+        self.assertIn(
+            "`implement-epic` → `implement-ticket` → `review-code-change`",
+            self.contract,
+        )
+        self.assertIn("retain task state", self.contract)
+        self.assertIn("poll or wait for asynchronous", self.contract)
+        self.assertIn(
+            "worker and subagent describe possible isolated execution roles",
+            self.contract,
+        )
+        self.assertIn("does not constrain the operating contract", self.contract)
+
     def test_implement_ticket_is_required_and_owns_one_ticket(self):
         self.assertIn("verify that `implement-ticket` is available", self.contract)
         self.assertIn(
@@ -138,6 +164,10 @@ def test_forward_cases_cover_composed_contract(self):
             "umbrella-closeout-separately",
             "unexpected-requires-epic-child-result",
             "parallel-nonoverlap-required",
+            "equivalent-isolated-context-profile",
+            "missing-review-dependency-through-ticket",
+            "missing-isolation-capability",
+            "missing-asynchronous-wait",
         }
         self.assertEqual(required, set(self.cases))
         self.assertEqual(required, set(self.results))
@@ -158,6 +188,16 @@ def test_forward_results_preserve_critical_boundaries(self):
             "blocked",
             self.results["unexpected-requires-epic-child-result"]["workflow_state"],
         )
+        self.assertEqual(
+            "epic_children_merged",
+            self.results["equivalent-isolated-context-profile"]["workflow_state"],
+        )
+        for case_id in (
+            "missing-review-dependency-through-ticket",
+            "missing-isolation-capability",
+            "missing-asynchronous-wait",
+        ):
+            self.assertEqual("blocked", self.results[case_id]["workflow_state"])
 
 
 if __name__ == "__main__":

skills/implement-ticket/SKILL.md

@@ -34,7 +34,8 @@ same-numbered issue from the PR host for the real tracker ticket.
 
 A compatible agentic runtime must be able to:
 
-- load this skill and repository-owned skill dependencies;
+- load `implement-ticket` and repository-owned `review-code-change` by stable
+  skill name or an equivalent repository-owned dependency mechanism;
 - read repository instructions, tracker state, and structured relationships;
 - inspect and create isolated branch/worktree state;
 - edit files, run commands, commit, push, and manage PRs when authorized;
@@ -46,7 +47,8 @@ A compatible agentic runtime must be able to:
 Stop with an explicit missing-capability result when an applicable capability is
 unavailable. Product-specific discovery metadata such as `agents/openai.yaml`
 may exist, but it does not constrain the operating contract or require a
-particular agent product.
+particular agent product. Terms such as worker and subagent describe possible
+isolated execution roles, not required product APIs.
 
 ## Resolve the operating contract
 
@@ -93,10 +95,11 @@ manual transition. State that consequence before publication or merge. Do not
 use automatic closing syntax when its effect conflicts with the resolved
 completion policy.
 
-For a merge-inclusive run, verify before implementation that
-`review-code-change` is available and readable. It is the only local
-adversarial-review dependency. Return `blocked` when it is unavailable; do not
-substitute a third-party skill, generic self-review, or unreviewed merge path.
+Before implementation can produce a reviewed PR, verify that
+`review-code-change` is available and readable by stable name or an equivalent
+repository-owned dependency mechanism. It is the only local adversarial-review
+dependency. Return `blocked` before mutation when it is unavailable; do not
+substitute a third-party skill, generic self-review, or unreviewed path.
 
 ## Establish source-of-truth precedence
 

skills/implement-ticket/evals/cases.json

@@ -4,7 +4,8 @@
     "request": "Implement GitHub issue G-21.",
     "tracker_state": "Open standalone PR-sized ticket with complete requirements, no blockers, and no existing implementation.",
     "authority": "ready_pr_only",
-    "capabilities": ["review-code-change", "isolated_worktree", "github_pr"]
+    "runtime_profile": "Repository-owned skills and named subagents are available.",
+    "capabilities": ["stable_skill_loading", "structured_tracker_relationships", "isolated_worktree", "command_execution", "github_pr", "named_read_only_reviewer", "asynchronous_wait", "thread_aware_feedback"]
   },
   {
     "id": "authorized-merge-cleanup",
@@ -93,5 +94,31 @@
     "tracker_state": "G-35 has children and requires whole-epic delivery.",
     "incoming_handoff": "implement-ticket:requires-epic:github:G-35",
     "capabilities": []
+  },
+  {
+    "id": "equivalent-isolated-context-profile",
+    "request": "Implement GitHub issue G-38 to a ready PR.",
+    "tracker_state": "G-38 is a ready standalone ticket with no competing implementation.",
+    "runtime_profile": "The runtime calls isolated units jobs rather than subagents, but provides an exclusive mutating job and a fresh read-only audit job with equivalent permissions and integrity checks.",
+    "capabilities": ["stable_skill_loading", "structured_tracker_relationships", "isolated_worktree", "command_execution", "github_pr", "equivalent_read_only_reviewer", "asynchronous_wait", "thread_aware_feedback"]
+  },
+  {
+    "id": "missing-review-code-change",
+    "request": "Implement GitHub issue G-39.",
+    "tracker_state": "G-39 is otherwise ready and independently shippable.",
+    "runtime_profile": "The runtime cannot load repository-owned review-code-change by stable name or an equivalent dependency mechanism."
+  },
+  {
+    "id": "missing-isolation-capability",
+    "request": "Implement GitHub issue G-40.",
+    "tracker_state": "G-40 is otherwise ready and independently shippable.",
+    "runtime_profile": "The runtime cannot create or verify an exclusively owned implementation branch/worktree context."
+  },
+  {
+    "id": "missing-asynchronous-wait",
+    "request": "Implement and merge GitHub issue G-41.",
+    "tracker_state": "G-41 is otherwise ready and requires asynchronous CI and review gates.",
+    "authority": "merge_after_gates",
+    "runtime_profile": "The runtime cannot poll, wait, or resume after asynchronous gates."
   }
 ]

skills/implement-ticket/evals/results.json

@@ -78,5 +78,25 @@
     "case_id": "repeated-epic-handoff",
     "terminal_state": "blocked",
     "required_actions": ["report routing cycle detected", "perform no mutation", "do not emit another requires_epic"]
+  },
+  {
+    "case_id": "equivalent-isolated-context-profile",
+    "terminal_state": "ready_pr",
+    "required_actions": ["accept equivalent isolated context terminology", "verify exclusive mutating ownership", "invoke fresh read-only review", "publish one PR"]
+  },
+  {
+    "case_id": "missing-review-code-change",
+    "terminal_state": "blocked",
+    "required_actions": ["name review-code-change as unavailable", "perform no implementation mutation", "do not substitute generic or third-party review"]
+  },
+  {
+    "case_id": "missing-isolation-capability",
+    "terminal_state": "blocked",
+    "required_actions": ["name exclusive implementation isolation as unavailable", "perform no implementation mutation", "preserve existing worktrees"]
+  },
+  {
+    "case_id": "missing-asynchronous-wait",
+    "terminal_state": "blocked",
+    "required_actions": ["name asynchronous wait as unavailable", "perform no implementation mutation", "do not pretend CI or review gates passed"]
   }
 ]

skills/implement-ticket/scripts/tests/test_implement_ticket_contract.py

@@ -30,6 +30,10 @@ def setUpClass(cls):
         cls.linear_compact = compact(cls.linear)
         cls.gates_compact = compact(cls.gates)
         cls.result_compact = compact(cls.result)
+        cls.eval_contract = compact(
+            read(SKILL_ROOT / "evals" / "cases.json")
+            + read(SKILL_ROOT / "evals" / "results.json")
+        )
         cls.all_contract = compact(
             cls.skill + cls.github + cls.linear + cls.gates + cls.result
         )
@@ -45,8 +49,23 @@ def setUpClass(cls):
     def test_frontmatter_and_product_neutral_contract(self):
         self.assertTrue(self.skill.startswith("---\nname: implement-ticket\n"))
         self.assertNotIn("Codex", self.all_contract)
+        self.assertNotIn("OpenAI", self.all_contract)
+        self.assertNotIn("Codex", self.eval_contract)
+        self.assertNotIn("OpenAI", self.eval_contract)
         self.assertNotIn("code-review-pro", self.all_contract)
         self.assertIn("compatible agentic runtime", self.skill)
+        self.assertIn(
+            "`implement-ticket` and repository-owned `review-code-change` by stable skill name",
+            self.skill_compact,
+        )
+        self.assertIn(
+            "worker and subagent describe possible isolated execution roles",
+            self.skill_compact,
+        )
+        self.assertIn(
+            "does not constrain the operating contract",
+            self.skill_compact,
+        )
 
     def test_scope_is_exactly_one_ticket(self):
         self.assertIn("one ticket per branch, worktree, and PR", self.skill_compact)
@@ -106,6 +125,7 @@ def test_authority_does_not_expand(self):
 
     def test_review_dependency_and_integrity_are_preserved(self):
         self.assertIn("only local adversarial-review dependency", self.all_contract)
+        self.assertIn("Return `blocked` before mutation", self.skill_compact)
         self.assertIn("Do not substitute another skill", self.gates_compact)
         self.assertIn(
             "fresh or minimally inherited read-only context", self.gates_compact
@@ -162,6 +182,10 @@ def test_forward_cases_cover_required_boundaries(self):
             "explicit-child-not-redirected",
             "missing-implement-epic",
             "repeated-epic-handoff",
+            "equivalent-isolated-context-profile",
+            "missing-review-code-change",
+            "missing-isolation-capability",
+            "missing-asynchronous-wait",
         }
         self.assertEqual(required, set(self.cases))
         self.assertEqual(required, set(self.results))
@@ -184,6 +208,16 @@ def test_forward_results_enforce_routing_and_authority(self):
         self.assertEqual(
             "blocked", self.results["repeated-epic-handoff"]["terminal_state"]
         )
+        self.assertEqual(
+            "ready_pr",
+            self.results["equivalent-isolated-context-profile"]["terminal_state"],
+        )
+        for case_id in (
+            "missing-review-code-change",
+            "missing-isolation-capability",
+            "missing-asynchronous-wait",
+        ):
+            self.assertEqual("blocked", self.results[case_id]["terminal_state"])
         self.assertIn(
             "do not merge",
             self.results["clean-local-review-remote-pending"]["required_actions"],