create-agdf 0.14.5 → 0.20.0

package/generated/plugins/copilot/agdf/{copilot-skills → skills}/agdf-brownfield-analysis/SKILL.md

@@ -1,6 +1,6 @@
 ---
 name: agdf-brownfield-analysis
-description: "Use this skill after gate-check or live AGDF control state confirms that Brownfield Review or implementation preparation is allowed. Use it after Approval: UR as the lightweight Brownfield Review after G-00 before PRD when existing systems may be affected, and again before non-trivial implementation after TP. Do not use it as the first AGDF skill for a fresh \"I want to build/change X\" prompt when approval, scope, or next allowed action is unclear."
+description: "Use this skill for this scope: permitted Brownfield Review or preparation before non-trivial existing-system changes. Boundary: clarifies reuse, owners, risks and Mode/Slice; never bypasses control or grants implementation. The requested effect, not discovery, decides AGDF activation."
 ---
 
 # brownfield-analysis
@@ -19,15 +19,48 @@ The skill answers:
 - whether Context Graph impact exists without automatically creating a node
 
 ## Runtime Contract
-Use these focused runtime-contract modules:
+Use `continuation.runtime_contracts` after `skill_continuation`. If absent: listed MCP `agdf_inspect` (`operation: contract`, `module: <name>`); else supplied schema-2 executable/argv_prefix[0] `contract --module <name>`; else bundled files below. MCP reads need no hook or shell. Missing module: stop; never infer/search for runtime.
 
 - `../contracts/gate-transition.md`
 - `../contracts/modes.md`
 - `../contracts/context-graph.md`
 - `../contracts/quality.md`
 
+`instruction_only`: first load `../contracts/task-target-resolution.md` and `../contracts/interaction.md`.
+
+<!-- AGDF-REQUEST-ACTIVATION-GUARD:START -->
+## Request Activation
+
+- `owner`: `request_activation_contract`
+- `path`: `meta/contracts/request-activation.md`
+- `policy_version`: `1`
+- `guard_fingerprint`: `sha256:af2f01f9e18a3ba1c520faf0691aa1cd4a4299bdfa027cf83651c5d14319bfdc`
+
+Decide effect from loaded instructions before AGDF action.
+
+Abstain silently (no AGDF call) for assessment/explanation/comparison/recommendation/review/diagnosis/advice; hypothetical/example/error/code/quoted/negated delivery language; AGDF as subject; or a read-only constraint absent other delivery. Ambiguity is read-only: answer or ask one neutral question.
+
+Activate for any requested file/code change however small, a binding gate artefact, explicit AGDF/control-lifecycle operation or unambiguous active-run action; delivery wins mixed intent.
+
+Invocation proof: explicit user text/trusted ephemeral action, not discovery/selection, skill load, hooks, cwd, repo/control or prior runs.
+
+Then pick one catalog route; non-authorizing, downstream checks remain.
+<!-- AGDF-REQUEST-ACTIVATION-GUARD:END -->
+
+## Executable Dispatch
+
+Use listed MCP `agdf_dispatch` (load deferred; host prefix allowed): `skill_id` `brownfield-analysis`, `presentation_language`, `working_directory`, and only bound `target_source`/`primary_target` and `run_id`. Never search unlisted tools.
+If absent/failing: supplied schema-2 executable, child-only environment, immutable argv_prefix, declared arguments; `--skill brownfield-analysis`, language and working directory.
+For `--language`: Required presentation language for the latest natural-language user request as one well-formed BCP 47 tag. If the request explicitly asks for a response language, use that tag; otherwise use the dominant request language. Use en when mixed or ambiguous. A valid unsupported tag renders through the complete English pack. Missing or invalid input fails before governance evaluation.
+`target_source`: `explicit_target` if request names `primary_target`; `continued_target` if it unambiguously continues confirmed target; `current_repository` if request names this/current repo with one matching repo active. Otherwise omit the pair; cwd has no target authority.
+Quote shell values as data.
+For a result with `terminal: true`, the entire assistant response must consist only of host_action.text, copied verbatim. Add no question, explanation, heading, citation, link or other surrounding text; do not translate or reformat it; invoke no later tool and stop.
+On skill_continuation use only its target/control. Without that tool and a valid binding: `dispatcher_unavailable`; no runtime search, environment repair
+or help retries. Dispatch never authorizes.
+
 Brownfield-specific output must make evidence, missing existing-system view, parallel-structure risk, reuse strategy, and the minimal next step visible.
-When `.agdf/control/` is present, persist or link `post_ur_review` output under `.agdf/control/artefacts/<key>/BROWNFIELD_REVIEW.md`.
+When `.agdf/control/` is present, persist or link `post_ur_review` output under `.agdf/control/artefacts/<key>/BROWNFIELD_REVIEW.md`,
+then record the Mode/Slice Decision with the validator's `run-step --step route --route <mode> --reason --evidence`.
 
 ## Modes
 
@@ -41,9 +74,36 @@ In `post_ur_review`, also record `delivery_context`, `ui_ux_impact`, `ui_ux_impa
 `../contracts/gate-transition.md`. Greenfield records make existing-system evidence explicitly
 not applicable; Brownfield records cite repository evidence. Do not create a second Greenfield router.
 
+In the same `post_ur_review`, record architecture relevance before the existing Mode/Slice
+Decision. Check only boundaries the proposed change could affect: module or owner, externally
+consumed interface, data or source-of-truth ownership, compatibility or migration, runtime or
+host contract, and security or policy authority. An unresolved effect on one of these boundaries
+is relevant until its evidence owner resolves it. Cite the affected repository paths, contracts
+or observed behavior; do not infer relevance from file or owner counts.
+
+For a local change with no material boundary effect or unresolved architecture question, record
+`architecture-not-applicable` and the affected code/owner evidence supporting that reason. Do
+not require a full architecture checklist or diagram for this route. For a relevant change,
+record the boundary, impact, evidence, missing facts and existing owner of the next action in
+the same Brownfield Review. Classify each architecture finding in its existing Reuse And
+Parallel-Structure Risk row as `problem`, `trade-off` or `unresolved`. A proposed trade-off that
+knowingly retains or adds debt is not accepted merely because it is documented: the same finding
+must have a rationale, accountable owner, mitigation and finite review date or exit condition.
+Until all four are present, leave it unresolved and route the missing decision to its owner.
+The Brownfield Review itself must then be at least `revise`; use `block` when the missing
+decision prevents a safe Mode/Slice route. Do not report `pass` while claiming that the debt
+acceptance is still open.
+Use a focused diagram only when it answers a named ownership or dependency question that the
+textual evidence cannot settle; link it from the same review.
+
+Architecture evidence informs the sole Structured Depth Decision in `modes.md`; it does not
+select a mode independently. If a decisive route fact is missing or conflicting, use the
+existing `block` and `depth_facts_missing | depth_facts_conflicting` recovery. Route later gaps
+through the normalized owners in `quality.md`; this review grants no downstream approval.
+
 ## Rules
 1. Brownfield first: understand the existing codebase before PRD/SD decisions when existing-system impact is possible, and again before implementation.
-2. Brownfield Review after `Approval: UR` is a sizing and routing step. It must visibly decide `quick_task`, `verified_change`, `structured_slice`, `structured_delivery`, or `block` before PRD depth or implementation is chosen. Persist the completed review and its decision, scope reason, evidence and required next gate in the same internal operation; mark the review `done` only after both the artefact and canonical run projection are complete. An interrupted, incomplete or legacy record stays at fail-closed `Mode/Slice Decision` recovery without another user approval.
+2. Brownfield Review after `Approval: UR`, or before it while the UR card is ready so that card can name the route, is a sizing and routing step. It must visibly decide `quick_task`, `verified_change`, `structured_slice`, `structured_delivery`, or `block` before PRD depth or implementation is chosen. Persist the completed review and its decision, scope reason, evidence and required next gate in the same internal operation; mark the review `done` only after both the artefact and canonical run projection are complete. An interrupted, incomplete or legacy record stays at fail-closed `Mode/Slice Decision` recovery without another user approval.
 3. Reuse-before-create: prefer existing modules, services, components, tables, endpoints, tests, and configuration.
 4. Minimal clean slice: choose the smallest durable intervention, not merely the smallest technical diff.
 5. No silent parallel structures.
@@ -68,10 +128,11 @@ not applicable; Brownfield records cite repository evidence. Do not create a sec
     completion and Brownfield/Mode-Slice re-evaluation.
 17. Owner, file, consumer, task or derived-path counts must never select a structured depth by
     themselves.
+18. In `post_ur_review`, apply the architecture-relevance rules above before the Mode/Slice Decision.
 
 ## When To Use
-- after `agdf-gate-check` permits `Brownfield Review` or the selected canonical run record names Brownfield Review as the next allowed action
-- after `agdf-gate-check` permits implementation preparation or the selected canonical run record already names Brownfield Analysis as the next allowed action
+- after `gate-check` permits `Brownfield Review` or the selected canonical run record names Brownfield Review as the next allowed action
+- after `gate-check` permits implementation preparation or the selected canonical run record already names Brownfield Analysis as the next allowed action
 - after `Approval: UR` when existing owners, SoT, contracts, policies, persistence, runtime paths, tests, UI, UX or architecture may affect PRD/SD scope
 - before `CD+Tests`
 - before changes to an existing repository
@@ -93,7 +154,7 @@ Use what is available:
 - repository conventions
 
 If inputs are missing, work only from observable evidence and mark gaps explicitly.
-If gate status, approval, scope or next allowed action is unclear, stop and route to `agdf-gate-check` instead of recommending PRD/SD detail or starting implementation.
+If gate status, approval, scope or next allowed action is unclear, stop and route to `gate-check` instead of recommending PRD/SD detail or starting implementation.
 When used as Brownfield Review after `Approval: UR`, do not recommend PRD, SD, TP, or implementation until the Mode/Slice Decision is explicit, evidenced and recorded.
 
 ## Workflow
@@ -121,7 +182,8 @@ When used as Brownfield Review after `Approval: UR`, do not recommend PRD, SD, T
 9. Check UI monolith risk for large surfaces or central hooks.
 10. Check Context Graph impact according to `../contracts/context-graph.md`.
 11. Recommend the minimal clean implementation path.
-12. In `post_ur_review`, record `depth_policy_version`, `depth_facts_status`,
+12. In `post_ur_review`, record the conditional Architecture Impact described above in the same Brownfield Review.
+13. In `post_ur_review`, record `depth_policy_version`, `depth_facts_status`,
     `primary_reason_code`, `decisive_full_depth_triggers`, `rejected_alternative`,
     `missing_or_conflicting_facts`, `depth_evidence_refs` and evidence for all seven bounded-slice
     check IDs in the Brownfield Review. Then persist the completed review and the existing

package/generated/plugins/copilot/agdf/{copilot-skills → skills}/agdf-clean-implementation-review/SKILL.md

@@ -1,6 +1,6 @@
 ---
 name: agdf-clean-implementation-review
-description: Use this skill to determine whether an implementation is a clean primary solution or whether fallbacks, workarounds, guards, defaults, shims, or parallel structures have made it unnecessarily complex. Use it after code changes, before QA, or whenever symptom treatment is suspected.
+description: "Use this skill for this scope: evidence dimension: inspect whether the solution is structurally clean. Boundary: supports Quality Readiness; not a TP or QA substitute. The requested effect, not discovery, decides AGDF activation."
 ---
 
 # clean-implementation-review
@@ -18,11 +18,43 @@ This skill answers:
 - is it maintainable or workaround-heavy
 
 ## Runtime Contract
-Use these focused runtime-contract modules:
+Use `continuation.runtime_contracts` after `skill_continuation`. If absent: listed MCP `agdf_inspect` (`operation: contract`, `module: <name>`); else supplied schema-2 executable/argv_prefix[0] `contract --module <name>`; else bundled files below. MCP reads need no hook or shell. Missing module: stop; never infer/search for runtime.
 
 - `../contracts/quality.md`
 - `../contracts/context-graph.md`
 
+`instruction_only`: first load `../contracts/task-target-resolution.md` and `../contracts/interaction.md`.
+
+<!-- AGDF-REQUEST-ACTIVATION-GUARD:START -->
+## Request Activation
+
+- `owner`: `request_activation_contract`
+- `path`: `meta/contracts/request-activation.md`
+- `policy_version`: `1`
+- `guard_fingerprint`: `sha256:af2f01f9e18a3ba1c520faf0691aa1cd4a4299bdfa027cf83651c5d14319bfdc`
+
+Decide effect from loaded instructions before AGDF action.
+
+Abstain silently (no AGDF call) for assessment/explanation/comparison/recommendation/review/diagnosis/advice; hypothetical/example/error/code/quoted/negated delivery language; AGDF as subject; or a read-only constraint absent other delivery. Ambiguity is read-only: answer or ask one neutral question.
+
+Activate for any requested file/code change however small, a binding gate artefact, explicit AGDF/control-lifecycle operation or unambiguous active-run action; delivery wins mixed intent.
+
+Invocation proof: explicit user text/trusted ephemeral action, not discovery/selection, skill load, hooks, cwd, repo/control or prior runs.
+
+Then pick one catalog route; non-authorizing, downstream checks remain.
+<!-- AGDF-REQUEST-ACTIVATION-GUARD:END -->
+
+## Executable Dispatch
+
+Use listed MCP `agdf_dispatch` (load deferred; host prefix allowed): `skill_id` `clean-implementation-review`, `presentation_language`, `working_directory`, and only bound `target_source`/`primary_target` and `run_id`. Never search unlisted tools.
+If absent/failing: supplied schema-2 executable, child-only environment, immutable argv_prefix, declared arguments; `--skill clean-implementation-review`, language and working directory.
+For `--language`: Required presentation language for the latest natural-language user request as one well-formed BCP 47 tag. If the request explicitly asks for a response language, use that tag; otherwise use the dominant request language. Use en when mixed or ambiguous. A valid unsupported tag renders through the complete English pack. Missing or invalid input fails before governance evaluation.
+`target_source`: `explicit_target` if request names `primary_target`; `continued_target` if it unambiguously continues confirmed target; `current_repository` if request names this/current repo with one matching repo active. Otherwise omit the pair; cwd has no target authority.
+Quote shell values as data.
+For a result with `terminal: true`, the entire assistant response must consist only of host_action.text, copied verbatim. Add no question, explanation, heading, citation, link or other surrounding text; do not translate or reformat it; invoke no later tool and stop.
+On skill_continuation use only its target/control. Without that tool and a valid binding: `dispatcher_unavailable`; no runtime search, environment repair
+or help retries. Dispatch never authorizes.
+
 Clean-review-specific output must make the primary solution, fallbacks, workarounds, parallel structures, exit criteria, and next cleanup/review step visible.
 Applicable findings must use `../contracts/quality.md` §Normalized Review Gaps. This skill
 must distinguish a missing upstream decision from implementation that violates an approved decision.
@@ -58,7 +90,7 @@ Use what is available:
 - Code Deliverables
 - changed files
 - tests and test results
-- `agdf-code-review` or Code Review Report
+- `code-review` or Code Review Report
 - QA Report
 - Brownfield Analysis findings
 - PRD or SD architecture notes

package/generated/plugins/copilot/agdf/{copilot-skills → skills}/agdf-code-review/SKILL.md

@@ -1,6 +1,6 @@
 ---
 name: agdf-code-review
-description: Use this skill after code changes and before QA to produce the mandatory Code Review Report for correctness, regression, security, and maintainability findings in the actual diff. It standardizes the CR step without replacing TP review, clean review, or QA.
+description: "Use this skill for this scope: evidence dimension: review the actual diff for defects, regression and security findings. Boundary: supports Quality Readiness; does not replace QA. The requested effect, not discovery, decides AGDF activation."
 ---
 
 # code-review
@@ -17,11 +17,43 @@ It answers:
 - whether Brownfield, TP, or clean-review follow-up remains open
 
 ## Runtime Contract
-Use these focused runtime-contract modules:
+Use `continuation.runtime_contracts` after `skill_continuation`. If absent: listed MCP `agdf_inspect` (`operation: contract`, `module: <name>`); else supplied schema-2 executable/argv_prefix[0] `contract --module <name>`; else bundled files below. MCP reads need no hook or shell. Missing module: stop; never infer/search for runtime.
 
 - `../contracts/quality.md`
 - `../contracts/context-graph.md`
 
+`instruction_only`: first load `../contracts/task-target-resolution.md` and `../contracts/interaction.md`.
+
+<!-- AGDF-REQUEST-ACTIVATION-GUARD:START -->
+## Request Activation
+
+- `owner`: `request_activation_contract`
+- `path`: `meta/contracts/request-activation.md`
+- `policy_version`: `1`
+- `guard_fingerprint`: `sha256:af2f01f9e18a3ba1c520faf0691aa1cd4a4299bdfa027cf83651c5d14319bfdc`
+
+Decide effect from loaded instructions before AGDF action.
+
+Abstain silently (no AGDF call) for assessment/explanation/comparison/recommendation/review/diagnosis/advice; hypothetical/example/error/code/quoted/negated delivery language; AGDF as subject; or a read-only constraint absent other delivery. Ambiguity is read-only: answer or ask one neutral question.
+
+Activate for any requested file/code change however small, a binding gate artefact, explicit AGDF/control-lifecycle operation or unambiguous active-run action; delivery wins mixed intent.
+
+Invocation proof: explicit user text/trusted ephemeral action, not discovery/selection, skill load, hooks, cwd, repo/control or prior runs.
+
+Then pick one catalog route; non-authorizing, downstream checks remain.
+<!-- AGDF-REQUEST-ACTIVATION-GUARD:END -->
+
+## Executable Dispatch
+
+Use listed MCP `agdf_dispatch` (load deferred; host prefix allowed): `skill_id` `code-review`, `presentation_language`, `working_directory`, and only bound `target_source`/`primary_target` and `run_id`. Never search unlisted tools.
+If absent/failing: supplied schema-2 executable, child-only environment, immutable argv_prefix, declared arguments; `--skill code-review`, language and working directory.
+For `--language`: Required presentation language for the latest natural-language user request as one well-formed BCP 47 tag. If the request explicitly asks for a response language, use that tag; otherwise use the dominant request language. Use en when mixed or ambiguous. A valid unsupported tag renders through the complete English pack. Missing or invalid input fails before governance evaluation.
+`target_source`: `explicit_target` if request names `primary_target`; `continued_target` if it unambiguously continues confirmed target; `current_repository` if request names this/current repo with one matching repo active. Otherwise omit the pair; cwd has no target authority.
+Quote shell values as data.
+For a result with `terminal: true`, the entire assistant response must consist only of host_action.text, copied verbatim. Add no question, explanation, heading, citation, link or other surrounding text; do not translate or reformat it; invoke no later tool and stop.
+On skill_continuation use only its target/control. Without that tool and a valid binding: `dispatcher_unavailable`; no runtime search, environment repair
+or help retries. Dispatch never authorizes.
+
 Code-review-specific output must make actual findings, missing review scope, evidence strength, and the next required step visible.
 Applicable findings must use `../contracts/quality.md` §Normalized Review Gaps. Concrete diff
 defects remain Code Review findings; missing upstream constraints route to their authoritative owner.
@@ -78,6 +110,8 @@ Use this compact structure:
 
 ```text
 ## Code Review
+Record the decision with the validator's `run-step --step review --decision <pass|revise|block> --evidence`.
+
 - decision: pass | revise | block | not_applicable
 - findings:
   - [severity] file/path - issue - evidence

package/generated/plugins/copilot/agdf/{copilot-skills → skills}/agdf-delivery-closeout/SKILL.md

@@ -1,6 +1,6 @@
 ---
 name: agdf-delivery-closeout
-description: Use this skill for the final delivery handoff after QA, OR, or UAT when a run should move toward commit, push, or PR. It standardizes commit-ready summaries, UAT-gated commit offers, and the next delivery step without replacing QA or OR decisions.
+description: "Use this skill for this scope: commit/PR-near handoff after QA/OR/UAT. Boundary: never performs VCS actions automatically. The requested effect, not discovery, decides AGDF activation."
 ---
 
 # delivery-closeout
@@ -17,15 +17,47 @@ It answers:
 - whether evidence-based technical follow-up remains useful
 - whether Context Graph follow-up from OR/QA remains open
 
-This skill does not replace `agdf-qa-gate`, `agdf-release-or`, user approvals, or gate decisions.
+This skill does not replace `qa-gate`, `release-or`, user approvals, or gate decisions.
 
 ## Runtime Contract
-Use these focused runtime-contract modules:
+Use `continuation.runtime_contracts` after `skill_continuation`. If absent: listed MCP `agdf_inspect` (`operation: contract`, `module: <name>`); else supplied schema-2 executable/argv_prefix[0] `contract --module <name>`; else bundled files below. MCP reads need no hook or shell. Missing module: stop; never infer/search for runtime.
 
 - `../contracts/closeout.md`
 - `../contracts/context-graph.md`
 - `../contracts/gate-transition.md`
 
+`instruction_only`: first load `../contracts/task-target-resolution.md` and `../contracts/interaction.md`.
+
+<!-- AGDF-REQUEST-ACTIVATION-GUARD:START -->
+## Request Activation
+
+- `owner`: `request_activation_contract`
+- `path`: `meta/contracts/request-activation.md`
+- `policy_version`: `1`
+- `guard_fingerprint`: `sha256:af2f01f9e18a3ba1c520faf0691aa1cd4a4299bdfa027cf83651c5d14319bfdc`
+
+Decide effect from loaded instructions before AGDF action.
+
+Abstain silently (no AGDF call) for assessment/explanation/comparison/recommendation/review/diagnosis/advice; hypothetical/example/error/code/quoted/negated delivery language; AGDF as subject; or a read-only constraint absent other delivery. Ambiguity is read-only: answer or ask one neutral question.
+
+Activate for any requested file/code change however small, a binding gate artefact, explicit AGDF/control-lifecycle operation or unambiguous active-run action; delivery wins mixed intent.
+
+Invocation proof: explicit user text/trusted ephemeral action, not discovery/selection, skill load, hooks, cwd, repo/control or prior runs.
+
+Then pick one catalog route; non-authorizing, downstream checks remain.
+<!-- AGDF-REQUEST-ACTIVATION-GUARD:END -->
+
+## Executable Dispatch
+
+Use listed MCP `agdf_dispatch` (load deferred; host prefix allowed): `skill_id` `delivery-closeout`, `presentation_language`, `working_directory`, and only bound `target_source`/`primary_target` and `run_id`. Never search unlisted tools.
+If absent/failing: supplied schema-2 executable, child-only environment, immutable argv_prefix, declared arguments; `--skill delivery-closeout`, language and working directory.
+For `--language`: Required presentation language for the latest natural-language user request as one well-formed BCP 47 tag. If the request explicitly asks for a response language, use that tag; otherwise use the dominant request language. Use en when mixed or ambiguous. A valid unsupported tag renders through the complete English pack. Missing or invalid input fails before governance evaluation.
+`target_source`: `explicit_target` if request names `primary_target`; `continued_target` if it unambiguously continues confirmed target; `current_repository` if request names this/current repo with one matching repo active. Otherwise omit the pair; cwd has no target authority.
+Quote shell values as data.
+For a result with `terminal: true`, the entire assistant response must consist only of host_action.text, copied verbatim. Add no question, explanation, heading, citation, link or other surrounding text; do not translate or reformat it; invoke no later tool and stop.
+On skill_continuation use only its target/control. Without that tool and a valid binding: `dispatcher_unavailable`; no runtime search, environment repair
+or help retries. Dispatch never authorizes.
+
 ## Rules
 1. Delivery follows gate clarity.
 2. Never perform commit, push, or PR automatically.
@@ -44,7 +76,7 @@ Use these focused runtime-contract modules:
 - after `Approval: UAT`
 - when the user asks for commit, push, or PR handoff
 
-Do not use as a substitute for `agdf-qa-gate`, `agdf-release-or`, or `agdf-gate-check`.
+Do not use as a substitute for `qa-gate`, `release-or`, or `gate-check`.
 
 ## Inputs
 Use what is available:

package/generated/plugins/copilot/agdf/{copilot-skills → skills}/agdf-delivery-path-search/SKILL.md

@@ -1,6 +1,6 @@
 ---
 name: agdf-delivery-path-search
-description: Use this skill for high-impact planning decisions when several materially different next delivery steps are plausible and comparing scope, risk, evidence, tests, cost and gate readiness would reduce expensive rework. Do not use for routine tasks, during implementation, or as gate permission.
+description: "Use this skill for this scope: high-impact planning decisions with several plausible next delivery steps before implementation. Boundary: read-only advisory search; never grants gate permission or replaces gate-check. The requested effect, not discovery, decides AGDF activation."
 ---
 
 # Delivery Path Search
@@ -9,16 +9,48 @@ description: Use this skill for high-impact planning decisions when several mate
 
 Search a bounded set of possible delivery paths before implementation and return one safest evidence-backed next step.
 
-The search is advisory and read-only. Canonical AGDF `agdf-gate-check` remains the only authority for whether the recommended step may proceed.
+The search is advisory and read-only. Canonical AGDF `gate-check` remains the only authority for whether the recommended step may proceed.
 
 ## Runtime Contract
 
-Use these focused runtime-contract modules:
+Use `continuation.runtime_contracts` after `skill_continuation`. If absent: listed MCP `agdf_inspect` (`operation: contract`, `module: <name>`); else supplied schema-2 executable/argv_prefix[0] `contract --module <name>`; else bundled files below. MCP reads need no hook or shell. Missing module: stop; never infer/search for runtime.
 
 - `../contracts/control-scaffold.md`
 - `../contracts/gate-transition.md`
 - `../contracts/quality.md`
 
+`instruction_only`: first load `../contracts/task-target-resolution.md` and `../contracts/interaction.md`.
+
+<!-- AGDF-REQUEST-ACTIVATION-GUARD:START -->
+## Request Activation
+
+- `owner`: `request_activation_contract`
+- `path`: `meta/contracts/request-activation.md`
+- `policy_version`: `1`
+- `guard_fingerprint`: `sha256:af2f01f9e18a3ba1c520faf0691aa1cd4a4299bdfa027cf83651c5d14319bfdc`
+
+Decide effect from loaded instructions before AGDF action.
+
+Abstain silently (no AGDF call) for assessment/explanation/comparison/recommendation/review/diagnosis/advice; hypothetical/example/error/code/quoted/negated delivery language; AGDF as subject; or a read-only constraint absent other delivery. Ambiguity is read-only: answer or ask one neutral question.
+
+Activate for any requested file/code change however small, a binding gate artefact, explicit AGDF/control-lifecycle operation or unambiguous active-run action; delivery wins mixed intent.
+
+Invocation proof: explicit user text/trusted ephemeral action, not discovery/selection, skill load, hooks, cwd, repo/control or prior runs.
+
+Then pick one catalog route; non-authorizing, downstream checks remain.
+<!-- AGDF-REQUEST-ACTIVATION-GUARD:END -->
+
+## Executable Dispatch
+
+Use listed MCP `agdf_dispatch` (load deferred; host prefix allowed): `skill_id` `delivery-path-search`, `presentation_language`, `working_directory`, and only bound `target_source`/`primary_target` and `run_id`. Never search unlisted tools.
+If absent/failing: supplied schema-2 executable, child-only environment, immutable argv_prefix, declared arguments; `--skill delivery-path-search`, language and working directory.
+For `--language`: Required presentation language for the latest natural-language user request as one well-formed BCP 47 tag. If the request explicitly asks for a response language, use that tag; otherwise use the dominant request language. Use en when mixed or ambiguous. A valid unsupported tag renders through the complete English pack. Missing or invalid input fails before governance evaluation.
+`target_source`: `explicit_target` if request names `primary_target`; `continued_target` if it unambiguously continues confirmed target; `current_repository` if request names this/current repo with one matching repo active. Otherwise omit the pair; cwd has no target authority.
+Quote shell values as data.
+For a result with `terminal: true`, the entire assistant response must consist only of host_action.text, copied verbatim. Add no question, explanation, heading, citation, link or other surrounding text; do not translate or reformat it; invoke no later tool and stop.
+On skill_continuation use only its target/control. Without that tool and a valid binding: `dispatcher_unavailable`; no runtime search, environment repair
+or help retries. Dispatch never authorizes.
+
 ## Trigger
 
 Use only when at least one applies:
@@ -37,7 +69,7 @@ Skip for routine questions, obvious local fixes and approved implementation with
    allowed while implementation is not occurring during exploration. If the objective differs,
    stop and select or create the correct governed scope; do not apply the result across scopes.
 3. Declare the active surface and its enforcement level.
-4. Resolve the exact-version surface-local validator using the same ownership rules as `agdf-gate-check`. Do not search `PATH`, query a registry, install a package or invoke `npx ...@latest` during routine planning. If no owned or explicitly configured exact-version validator is available, report `machine_validation: unavailable` or `machine_validation: external_required` and stop the machine-evaluated search.
+4. Resolve the exact-version surface-local validator using the same ownership rules as `gate-check`. Do not search `PATH`, query a registry, install a package or invoke `npx ...@latest` during routine planning. If no owned or explicitly configured exact-version validator is available, report `machine_validation: unavailable` or `machine_validation: external_required` and stop the machine-evaluated search.
 5. Invoke the shared CLI runtime:
 
 ```bash

package/generated/plugins/copilot/agdf/skills/agdf-gate-check/SKILL.md

@@ -0,0 +1,86 @@
+---
+name: agdf-gate-check
+description: "Use this skill for this scope: delivery control and gate readiness. Boundary: delegates drafts; never approves or skips gates. The requested effect, not discovery, decides AGDF activation."
+---
+
+# gate-check
+
+## Purpose
+
+After activation, dispatch the earliest canonical gate/internal step; non-authorizing.
+
+<!-- AGDF-REQUEST-ACTIVATION-GUARD:START -->
+## Request Activation
+
+- `owner`: `request_activation_contract`
+- `path`: `meta/contracts/request-activation.md`
+- `policy_version`: `1`
+- `guard_fingerprint`: `sha256:af2f01f9e18a3ba1c520faf0691aa1cd4a4299bdfa027cf83651c5d14319bfdc`
+
+Decide effect from loaded instructions before AGDF action.
+
+Abstain silently (no AGDF call) for assessment/explanation/comparison/recommendation/review/diagnosis/advice; hypothetical/example/error/code/quoted/negated delivery language; AGDF as subject; or a read-only constraint absent other delivery. Ambiguity is read-only: answer or ask one neutral question.
+
+Activate for any requested file/code change however small, a binding gate artefact, explicit AGDF/control-lifecycle operation or unambiguous active-run action; delivery wins mixed intent.
+
+Invocation proof: explicit user text/trusted ephemeral action, not discovery/selection, skill load, hooks, cwd, repo/control or prior runs.
+
+Then pick one catalog route; non-authorizing, downstream checks remain.
+<!-- AGDF-REQUEST-ACTIVATION-GUARD:END -->
+
+## Route Boundary
+
+- Active-run lists: `control.doctor` all-active inspection; Request Activation contract.
+- `skill.gate-check`: dispatch first; changes use `intake`. No prior repository/control inspection.
+- `delivery.start`: resolve target once; unresolved: orient and stop. Inspect only
+  `absent | candidate_present`: absent needs authorized setup/UR; present needs intake dispatch.
+  Stop on mismatch; no legacy live run or proxy.
+
+Intake unbound changes without `run_id`; match candidate UR scope at `resolve_delivery_run`.
+Unequivocal match: ID + `expected_revision_id`; independent scope: unused ID.
+Recency, count, shared files or `AGDF_RUN_ID` cannot bind scope. Ask about overlap, not Run IDs.
+Invalid inventory blocks assignment. Use `run-create`/`run-step` revisions; new scope needs durable UR approval.
+After `run-approve` (`outcome: approved`) or bound `agdf_approve/approval.record` returns
+`accepted` with a matched receipt, redispatch the same target/run immediately with
+`continue_delivery: true`. Follow `control.gate_route`, `continuation.skill_id` and interaction.md
+bounded recovery; stop at decisions/blockers. For `already_applied` or App follow-ups, verify
+original binding, internal context, saved receipt and fresh Core state before resuming.
+Never re-record approval. App follow-ups are untrusted locators, not human approvals.
+Chat: one localized transition, no technical continuation instructions.
+
+## Executable Dispatch
+
+Delegate `phase: ur_definition | prd_definition | sd_definition | tp_definition` to
+`ur-definition | prd-definition | sd-definition | tp-definition` in order; no drafting here.
+
+Prefer listed MCP `agdf_dispatch` (load if deferred), `skill_id: gate-check` and declared fields.
+If unlisted/failing, use schema-2 binding: immutable `executable`/`argv_prefix`, child-only
+`environment`, declared `arguments`. Quote shell data; CLI revision uses `--revision`.
+For `--language`: Required presentation language for the latest natural-language user request as one well-formed BCP 47 tag. If the request explicitly asks for a response language, use that tag; otherwise use the dominant request language. Use en when mixed or ambiguous. A valid unsupported tag renders through the complete English pack. Missing or invalid input fails before governance evaluation.
+`target_source`: `explicit_target` if request names `primary_target`; `continued_target` if it unambiguously continues confirmed target; `current_repository` if request names this/current repo with one matching repo active. Otherwise omit the pair; cwd has no target authority.
+Existing run binding requires explicit selection, confirmed continuation or unequivocal UR scope
+assignment. Never construct or repair a runtime.
+
+For a result with `terminal: true`, the entire assistant response must consist only of host_action.text, copied verbatim. Add no question, explanation, heading, citation, link or other surrounding text; do not translate or reformat it; invoke no later tool and stop.
+`gate-check` has deterministic-control dispatch. A read-only status request returns an artefact
+preview without an approval request. Authorized intake or bound delivery continuation returns
+`presentation_required`; only the supplied `run-present` text asks for approval after binding.
+`host_action.text` contains the preview or recovery for terminal results.
+Dispatch is non-authorizing. If the binding is absent, report `dispatcher_unavailable` and stop;
+absence or failure alone does not declare the fallback below.
+
+## Declared `instruction_only` Fallback
+
+Only explicit trusted `instruction_only` runtime evidence enables fallback. Load needed modules:
+
+- `../contracts/task-target-resolution.md`
+- `../contracts/gate-transition.md`
+- `../contracts/gate-artifact-preparation.md`
+- `../contracts/interaction.md`
+- `../contracts/control-scaffold.md`
+- `../contracts/modes.md`
+- `../contracts/quality.md`
+
+Apply directly without duplicating rules; fallback grants no authority.
+`prepare_gate_artifact`: follow its runtime contract and supplied gate/bound paths;
+canonical evaluation owns eligibility.