create-agdf 0.2.2 → 0.2.4

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Files changed (36) hide show
  1. package/README.md +6 -0
  2. package/bin/create-agdf.js +522 -36
  3. package/generated/.agdf/control/README.md +17 -1
  4. package/generated/.agdf/control/templates/AGDF_RUN.md +23 -1
  5. package/generated/.agdf/control/templates/MASTER_BACKLOG.md +16 -11
  6. package/generated/.agdf/control/templates/artefacts/OR.md +86 -0
  7. package/generated/.agdf/control/templates/artefacts/PRD.md +42 -0
  8. package/generated/.agdf/control/templates/artefacts/QA_REPORT.md +38 -0
  9. package/generated/.agdf/control/templates/artefacts/SD.md +42 -0
  10. package/generated/.agdf/control/templates/artefacts/TP.md +36 -0
  11. package/generated/.agdf/control/templates/artefacts/UR.md +41 -0
  12. package/generated/.github/copilot-instructions.md +3 -0
  13. package/generated/.github/instructions/agdf-governance.instructions.md +1 -0
  14. package/generated/.github/skills/agdf-brownfield-analysis/SKILL.md +24 -17
  15. package/generated/.github/skills/agdf-gate-check/SKILL.md +53 -5
  16. package/generated/.github/skills/agdf-release-or/SKILL.md +13 -9
  17. package/generated/.github/skills/agdf-runtime-contract.md +60 -2
  18. package/generated/AGENTS.md +15 -2
  19. package/generated/plugins/agdf/.codex-plugin/plugin.json +1 -1
  20. package/generated/plugins/agdf/control/README.md +17 -1
  21. package/generated/plugins/agdf/control/templates/AGDF_RUN.md +23 -1
  22. package/generated/plugins/agdf/control/templates/MASTER_BACKLOG.md +16 -11
  23. package/generated/plugins/agdf/control/templates/artefacts/OR.md +86 -0
  24. package/generated/plugins/agdf/control/templates/artefacts/PRD.md +42 -0
  25. package/generated/plugins/agdf/control/templates/artefacts/QA_REPORT.md +38 -0
  26. package/generated/plugins/agdf/control/templates/artefacts/SD.md +42 -0
  27. package/generated/plugins/agdf/control/templates/artefacts/TP.md +36 -0
  28. package/generated/plugins/agdf/control/templates/artefacts/UR.md +41 -0
  29. package/generated/plugins/agdf/meta/agdf-agent-router.md +15 -2
  30. package/generated/plugins/agdf/meta/agdf-constitution.md +22 -5
  31. package/generated/plugins/agdf/meta/agdf-plugin.definition.json +4 -4
  32. package/generated/plugins/agdf/meta/agdf-runtime-contract.md +60 -2
  33. package/generated/plugins/agdf/skills/brownfield-analysis/SKILL.md +24 -17
  34. package/generated/plugins/agdf/skills/gate-check/SKILL.md +53 -5
  35. package/generated/plugins/agdf/skills/release-or/SKILL.md +13 -9
  36. package/package.json +1 -1
@@ -28,6 +28,7 @@ It reports:
28
28
  Use `../agdf-runtime-contract.md` for Quality Contract output, Context Graph fields, gate terms, and non-duplication rules.
29
29
 
30
30
  OR-specific output must make gate status, delivered and intentionally not delivered content, missing approvals, missing evidence, risks, retained fallbacks, and the next permissible step visible.
31
+ When `.agdf/control/` is present, persist or link the OR under `.agdf/control/artefacts/<key>/OR.md` and reference it from `AGDF_RUN.md` or `MASTER_BACKLOG.md` when it is steering-relevant.
31
32
 
32
33
  ## Rules
33
34
  1. OR is always allowed and mandatory for relevant runs.
@@ -77,15 +78,16 @@ If information is missing, state the gap instead of guessing.
77
78
  2. Determine report depth:
78
79
  - `OR-lite`: gate status, allowed/forbidden outputs, missing approval, next step
79
80
  - `OR-full`: also delivered artefacts, TP coverage, Brownfield fit, solution integrity, risks, fallbacks, QA/UAT status
80
- 3. Record delivered vs intentionally not delivered content.
81
- 4. Summarize TP coverage if a TP exists.
82
- 5. Summarize Brownfield fit and solution integrity if reviewed.
83
- 6. Summarize tests and verification.
84
- 7. Summarize documentation and Context Graph impact if relevant.
85
- 8. Name retained fallbacks and exit criteria.
86
- 9. Set exactly one next permissible step.
87
- 10. Set exactly one quality outlook.
88
- 11. State whether `agdf-delivery-closeout` is the next operational handoff step when code changes exist and the delivery state allows it.
81
+ 3. Record the selected `Report mode` and durable OR target or link when `.agdf/control/` exists.
82
+ 4. Record delivered vs intentionally not delivered content.
83
+ 5. Summarize TP coverage if a TP exists.
84
+ 6. Summarize Brownfield fit and solution integrity if reviewed.
85
+ 7. Summarize tests and verification.
86
+ 8. Summarize documentation and Context Graph impact if relevant.
87
+ 9. Name retained fallbacks and exit criteria.
88
+ 10. Set exactly one next permissible step.
89
+ 11. Set exactly one quality outlook.
90
+ 12. State whether `agdf-delivery-closeout` is the next operational handoff step when code changes exist and the delivery state allows it.
89
91
 
90
92
  ## Output
91
93
  Use a compact structure:
@@ -93,6 +95,8 @@ Use a compact structure:
93
95
  ```text
94
96
  ## OR
95
97
  - gate:
98
+ - report_mode:
99
+ - artefact:
96
100
  - status:
97
101
  - delivered:
98
102
  - intentionally_not_delivered:
@@ -12,21 +12,54 @@ It is the operational reference for skill outputs, but it is not a second copy o
12
12
 
13
13
  Quick Task Mode must not become ritual gate overhead.
14
14
  Structured Delivery must not bypass missing approvals.
15
+ New product semantics, functional change, user-visible behaviour, policy, persistence, architecture or release-critical work requires a durable UR in `.agdf/control/` or a linked authoritative repository SoT before later artefacts or implementation.
16
+ UR, PRD, SD, TP and QA report approvals require durable artefacts or linked authoritative repository SoT entries before the next gate can open.
15
17
 
16
18
  ## Gate Rules
17
19
 
18
20
  - User gates: `UR -> PRD -> SD -> TP -> QA -> UAT`
19
- - Internal mandatory steps: `Brownfield Analysis -> CD+Tests -> CR -> OR`
21
+ - Internal mandatory steps: `Brownfield Review -> Mode/Slice Decision -> Brownfield Analysis -> CD+Tests -> CR -> OR`
20
22
  - Exact approval formula: `Approval: <GateName>`
21
23
  - Legacy alias: `Freigabe: <GateName>` may be accepted when reviewing older German runs, but all new outputs must use `Approval: <GateName>`.
22
- - Implicit consent is not approval.
24
+ - Implicit consent is not approval. Phrases such as "ok", "go ahead", "do it", "approved", "continue", "leg los" or "looks good" are work intent, not gate approval.
23
25
  - The earliest blocking gate wins.
26
+ - For UR, PRD, SD, TP and QA, approval and durable artefact presence are separate requirements. A gate is not satisfied by approval text alone when its artefact is missing from `.agdf/control/` or not linked to the authoritative repository SoT.
27
+ - Approval of one user gate permits work on the next allowed gate artefact or required internal step only. It does not permit implementation unless the approved gate is `TP` and required internal implementation prerequisites are satisfied.
28
+ - Gates must never be skipped or inferred from urgency, tone, chat history, task wording or an instruction to "start". The next allowed action is always the next unsatisfied gate or internal mandatory step.
29
+ - `Approval: UR` permits Brownfield Review after G-00 first, then a Mode/Slice Decision. It never permits implementation by itself.
30
+ - PRD, SD and TP depth is chosen after Brownfield Review through the Mode/Slice Decision, not before existing-system impact is understood.
31
+ - `Approval: PRD` permits Solution Design drafting, not implementation.
32
+ - `Approval: SD` permits Task/Test Plan drafting, not implementation.
33
+ - `Approval: TP` permits implementation-preparation Brownfield Analysis and then CD+Tests when Brownfield evidence supports it.
24
34
  - `CD+Tests` is implementation and test status only, not a delivery signal.
25
35
  - `agdf-code-review` standardizes the mandatory `CR` step, but does not replace QA.
26
36
  - `agdf-qa-gate` decides final `pass | revise | block`.
27
37
  - `agdf-release-or` reports the run, but does not replace QA or produce the operative commit/PR handoff.
28
38
  - `agdf-delivery-closeout` owns the operative Git handoff text when QA/OR/UAT state allows it.
29
39
 
40
+ ## Brownfield Review After G-00
41
+
42
+ After an approved and durable UR, perform a lightweight `Brownfield Review` before PRD when Brownfield, UI, UX, admin, runtime, policy, persistence, architecture, ownership or source-of-truth impact is possible.
43
+
44
+ This review is not a new user approval gate, not PRD, not SD, not TP and not implementation permission.
45
+ It is the sizing and routing step that decides how much later gate discipline is actually needed.
46
+
47
+ Its output is limited to:
48
+
49
+ - workstream or follow-up-slice classification
50
+ - `Mode/Slice Decision`: `quick_task | structured_slice | structured_delivery | block`
51
+ - affected existing owners, source-of-truth artefacts, contracts, policies, runtime paths or tests
52
+ - reuse-before-create risks and parallel-structure risks
53
+ - open questions that PRD or SD must resolve
54
+ - whether implementation-preparation Brownfield Analysis will also be needed after TP
55
+
56
+ Mode/Slice Decision rules:
57
+
58
+ - `quick_task`: use only when Brownfield Review shows a narrow local change, no new product semantics beyond the approved UR, no architecture/policy/persistence/contract expansion, and evidence is sufficient to proceed with a small implementation plus relevant checks.
59
+ - `structured_slice`: use when some formal artefacts are needed, but they can stay intentionally small and scoped to the approved slice.
60
+ - `structured_delivery`: use when the change has broad product, architecture, runtime, policy, persistence, release or cross-owner impact.
61
+ - `block`: use when ownership, SoT, impact, evidence or product direction is not clear enough to choose a safe path.
62
+
30
63
  ## Quality Contract Output
31
64
 
32
65
  Structured Delivery skills use this shape when relevant:
@@ -63,9 +96,34 @@ When a repository needs durable AGDF state, use the plugin-local `control/` scaf
63
96
  - `SOT_REGISTRY.md` prevents parallel sources of truth.
64
97
  - `CONTEXT_GRAPH.md` stores durable Brownfield findings, decisions, risks, evidence and exit criteria.
65
98
  - `AGENT_QUALITY_CONTRACTS.json` stores reusable block, revise and warning conditions.
99
+ - OR reports live under `.agdf/control/artefacts/<key>/OR.md` when a run closeout is steering-relevant or should be auditable beyond the chat.
66
100
 
67
101
  The scaffold is not a second documentation site. Link to authoritative artefacts instead of copying them.
68
102
 
103
+ ## Native Runtime
104
+
105
+ AGDF must work natively through the plugin router, skills and live `.agdf/control/` artefacts.
106
+ Agents should read the live control state and apply the gate model directly before reaching for helper commands.
107
+
108
+ Machine-readable commands such as `doctor --json`, `gate-check --json` and `delivery-map --json` are validators and automation interfaces.
109
+ They make the repository state checkable, but they do not replace the native skill workflow or the durable control artefacts.
110
+
111
+ ## Delivery Map
112
+
113
+ `delivery-map --json` is the machine-readable delivery picture for the active control state.
114
+ It derives, but does not replace, the live `.agdf/control/AGDF_RUN.md` and `MASTER_BACKLOG.md` state.
115
+
116
+ It must expose:
117
+
118
+ - approved or active artefacts and approvals
119
+ - `UR -> PRD -> SD -> TP -> QA_REPORT` relationships
120
+ - evidence refs, missing evidence and declared risks
121
+ - Context Graph impact and gate effect
122
+ - findings that explain why the delivery picture is `pass | warn | revise | block`
123
+
124
+ Missing relationship evidence in the Artefact Chain is at least `revise` once the related gate artefact is approved or passed.
125
+ Declared missing evidence, declared risk or Context Graph gate effect may escalate the delivery map to `warn`, `revise` or `block`.
126
+
69
127
  ## Skill Output
70
128
 
71
129
  Every skill output should be as short as possible and as concrete as needed.
@@ -29,17 +29,26 @@ Use Quick Task Mode for questions, small reviews, local debugging and narrow fix
29
29
 
30
30
  Use Structured Delivery Mode for new capabilities, architecture, policy, persistence, release-critical work, formal artefacts or explicit approvals.
31
31
 
32
+ Any new product semantics, functional change or user-visible behaviour change requires a durable UR first.
33
+ Persist it in `.agdf/control/` or link it from there to the target repository's authoritative source of truth before PRD, SD, TP, Brownfield Analysis or implementation.
34
+ Treat approval text and durable artefact presence as separate checks for UR, PRD, SD, TP and QA report decisions.
35
+ Do not infer approval from "ok", "go ahead", "do it", "approved", "continue", "leg los" or similar wording.
36
+ Those phrases may express intent to proceed, but they do not unlock a gate unless the exact gate formula is present.
37
+ After `Approval: UR`, run lightweight Brownfield Review before PRD when Brownfield, ownership, runtime, policy, persistence, architecture, UI or UX impact is possible.
38
+ Brownfield Review must produce a Mode/Slice Decision: `quick_task`, `structured_slice`, `structured_delivery` or `block`.
39
+ Do not assume the full PRD/SD/TP chain before that decision; use only as much gate depth as the reviewed change size justifies.
40
+
32
41
  Quick Task Mode must still cite evidence and close with a concrete result.
33
42
  Structured Delivery Mode must respect gates, reviews and closeout discipline.
34
43
 
35
44
  ## Skill Routing
36
45
  | Skill | Use For | Boundary |
37
46
  |---|---|---|
38
- | `agdf-brownfield-analysis` | after gate-check permits implementation preparation, before non-trivial changes in existing systems | clarifies reuse, owners and risks; never bypasses gate-check |
47
+ | `agdf-brownfield-analysis` | after gate-check permits Brownfield Review or implementation preparation, before non-trivial changes in existing systems | clarifies reuse, owners, risks and Mode/Slice Decision; never bypasses gate-check; Brownfield Review is not implementation permission |
39
48
  | `agdf-clean-implementation-review` | inspect workarounds, fallbacks, parallel structures or symptom fixes | not a TP or QA substitute |
40
49
  | `agdf-code-review` | mandatory CR step after code changes, focused on defects, regression and security findings | does not replace TP review, clean review or QA |
41
50
  | `agdf-delivery-closeout` | commit/PR-near handoff after QA/OR/UAT | never performs VCS actions automatically |
42
- | `agdf-gate-check` | new build/change intent, unclear approval, Structured Delivery, later-gate artefact requested | does not create later artefacts |
51
+ | `agdf-gate-check` | new build/change intent, unclear approval, Structured Delivery, later-gate artefact requested | does not create later artefacts or skip Mode/Slice Decision after Brownfield Review |
43
52
  | `agdf-qa-gate` | final QA decision | only instance for `pass | revise | block` |
44
53
  | `agdf-release-or` | auditable closeout for every relevant run | not a QA substitute |
45
54
  | `agdf-task-plan-review` | after implementation and before QA, verify TP coverage | no final QA decision |
@@ -47,6 +56,7 @@ Structured Delivery Mode must respect gates, reviews and closeout discipline.
47
56
  Select exactly one primary skill first.
48
57
  Add more only when they cover a distinct concrete risk dimension.
49
58
  Do not choose `agdf-brownfield-analysis` as the first primary skill for a fresh "I want to build/change X" prompt unless `agdf-gate-check` or existing live AGDF control state already makes implementation preparation the next allowed action.
59
+ Never jump directly from `Approval: UR`, implicit consent, or a generic "start" request to implementation. Route to Brownfield Review, then Mode/Slice Decision, then the smallest safe next gate or Quick Task execution.
50
60
 
51
61
  ## Runtime Contract
52
62
  For repeated output, gate, Quality Contract and Context Graph rules, use:
@@ -61,6 +71,9 @@ Use templates as starting points only.
61
71
 
62
72
  Do not let chat history become the source of truth for gate state, approvals, evidence, backlog status or durable Brownfield knowledge.
63
73
 
74
+ AGDF is a native agent workflow first: read the live control state, apply the active skill and make the next allowed step explicit.
75
+ Use `doctor`, `gate-check --json` or `delivery-map --json` as validators when a machine-readable proof is useful, not as a substitute for the router or skills.
76
+
64
77
  ## Closeout
65
78
  After relevant code, documentation, skill or governance changes:
66
79
 
@@ -1,6 +1,6 @@
1
1
  {
2
2
  "name": "agdf",
3
- "version": "0.2.2",
3
+ "version": "0.2.4",
4
4
  "description": "Codex-first workflow skills for gated, auditable AI-assisted software delivery.",
5
5
  "author": {
6
6
  "name": "AI-native Governance & Delivery Framework",
@@ -9,7 +9,13 @@ Use this scaffold when a target repository should keep AGDF state outside chat h
9
9
  | File | Purpose |
10
10
  |---|---|
11
11
  | `templates/AGDF_RUN.md` | Template for current run state: mode, gate, approvals, evidence, risks, next allowed action |
12
- | `templates/MASTER_BACKLOG.md` | Template for the living backlog pointer: active initiatives and current UR/PRD/SD/TP artefacts |
12
+ | `templates/MASTER_BACKLOG.md` | Template for the living backlog pointer: active initiatives and current UR/PRD/SD/TP/QA/OR artefacts |
13
+ | `templates/artefacts/UR.md` | Template for the durable user requirement artefact of a work item |
14
+ | `templates/artefacts/PRD.md` | Template for the durable product requirements artefact |
15
+ | `templates/artefacts/SD.md` | Template for the durable solution design artefact |
16
+ | `templates/artefacts/TP.md` | Template for the durable task and test plan artefact |
17
+ | `templates/artefacts/QA_REPORT.md` | Template for the durable QA decision report |
18
+ | `templates/artefacts/OR.md` | Template for the durable Orchestration Report closeout |
13
19
  | `templates/SOT_REGISTRY.md` | Template for the source-of-truth registry: one primary owner per domain |
14
20
  | `templates/CONTEXT_GRAPH.md` | Template for project memory: durable decisions, Brownfield findings, risks, evidence and exit criteria |
15
21
  | `templates/AGENT_QUALITY_CONTRACTS.json` | Template for machine-readable warning, revise and block contracts for review and QA |
@@ -21,6 +27,14 @@ Use this scaffold when a target repository should keep AGDF state outside chat h
21
27
  control/
22
28
  AGDF_RUN.md
23
29
  MASTER_BACKLOG.md
30
+ artefacts/
31
+ <work-item>/
32
+ UR.md
33
+ PRD.md
34
+ SD.md
35
+ TP.md
36
+ QA_REPORT.md
37
+ OR.md
24
38
  SOT_REGISTRY.md
25
39
  CONTEXT_GRAPH.md
26
40
  AGENT_QUALITY_CONTRACTS.json
@@ -49,6 +63,8 @@ npm create agdf@latest gate-check --json
49
63
 
50
64
  - `AGDF_RUN.md` is the current run dashboard.
51
65
  - `MASTER_BACKLOG.md` points to active delivery work; detailed artefacts live beside the work item.
66
+ - UR, PRD, SD, TP and QA decisions require durable artefacts or links to the authoritative repository source of truth before the next gate can open.
67
+ - OR records the run closeout and next permissible step; it does not approve later gates.
52
68
  - `SOT_REGISTRY.md` decides which document owns which domain.
53
69
  - `CONTEXT_GRAPH.md` records durable project knowledge only when it has evidence and an exit criterion.
54
70
  - `AGENT_QUALITY_CONTRACTS.json` names reusable block, revise and warning conditions.
@@ -5,7 +5,7 @@
5
5
  - run_id:
6
6
  - started_at:
7
7
  - mode: `quick_task | structured_delivery`
8
- - current_gate: `none | UR | PRD | SD | TP | CD+Tests | CR | QA | UAT | OR`
8
+ - current_gate: `none | UR | Brownfield Review | Mode/Slice Decision | PRD | SD | TP | Quick Task Execution | Brownfield Analysis | CD+Tests | CR | QA | UAT | OR`
9
9
  - decision: `pass | revise | block | in_progress`
10
10
  - owner:
11
11
 
@@ -44,10 +44,32 @@ Valid approval format for new runs: `Approval: <GateName>`.
44
44
  | PRD | | `draft | approved | superseded | not_applicable` | |
45
45
  | SD | | `draft | approved | superseded | not_applicable` | |
46
46
  | TP | | `draft | approved | superseded | not_applicable` | |
47
+ | Brownfield Review | | `missing | done | not_applicable` | |
47
48
  | Review | | `missing | done | not_applicable` | |
48
49
  | QA | | `missing | pass | revise | block | not_applicable` | |
49
50
  | OR | | `missing | done | not_applicable` | |
50
51
 
52
+ ## Mode / Slice Decision
53
+
54
+ Set this after Brownfield Review. Do not assume the full gate chain before the existing-system impact is understood.
55
+
56
+ - decision: `undecided | quick_task | structured_slice | structured_delivery | block`
57
+ - required_next_gate: `none | PRD | SD | TP | Brownfield Analysis`
58
+ - scope_reason:
59
+ - evidence:
60
+
61
+ ## Artefact Chain
62
+
63
+ Keep the active work item traceable. A gate may open only when the previous gate has both exact approval and a durable or linked artefact.
64
+
65
+ | From | Relationship | To | Evidence |
66
+ |---|---|---|---|
67
+ | UR | `approved_by` | `Approval: UR` | |
68
+ | PRD | `derived_from` | UR | |
69
+ | SD | `derived_from` | PRD | |
70
+ | TP | `derived_from` | SD | |
71
+ | QA_REPORT | `tests` | TP | |
72
+
51
73
  ## Evidence
52
74
 
53
75
  | Evidence | Source | Covers | Strength |
@@ -4,23 +4,28 @@ This file is the living control pointer for active AGDF delivery work. Detailed
4
4
 
5
5
  ## Rules
6
6
 
7
- 1. Keep only steering-relevant work in `Active Backlog`.
8
- 2. Every active item has exactly one current spec pointer.
9
- 3. Historical artefacts remain linkable but do not override the current pointer.
10
- 4. If an item is superseded, mark the replacement.
11
- 5. If active work grows beyond roughly ten items, triage the backlog.
7
+ 1. New product semantics, functional change or user-visible behaviour change starts as a new UR draft in a stable local artefact path, for example `.agdf/control/artefacts/<key>/UR.md`, unless an authoritative repository SoT already exists and is linked here.
8
+ 2. `Approval: UR` adds or updates exactly one steering row with a link to the UR.
9
+ 3. PRD, SD, TP and QA report artefacts are created for the same work item after the previous gate is approved and persisted or linked.
10
+ 4. `Approval: PRD`, `Approval: SD`, `Approval: TP` and `Approval: QA` require the corresponding durable artefact or linked repository SoT before the next gate can open.
11
+ 5. Generic consent such as "ok", "go ahead", "do it", "continue", "leg los" or "approved" does not update a gate. Only `Approval: <GateName>` does.
12
+ 6. Keep only steering-relevant work in `Active Backlog`.
13
+ 7. Every active item has exactly one current spec pointer and a visible artefact chain.
14
+ 8. Historical artefacts remain linkable but do not override the current pointer.
15
+ 9. If an item is superseded, mark the replacement.
16
+ 10. If active work grows beyond roughly ten items, triage the backlog.
12
17
 
13
18
  ## Active Backlog
14
19
 
15
- | Prio | Key | Title | Status | Current Spec | Notes |
16
- |---:|---|---|---|---|---|
17
- | P1 | | | `in_progress | blocked` | | |
20
+ | Prio | Key | Title | Status | UR | PRD | SD | TP | QA | OR | Current Spec | Notes |
21
+ |---:|---|---|---|---|---|---|---|---|---|---|---|
22
+ | P1 | | | `in_progress | blocked` | | | | | | | | |
18
23
 
19
24
  ## Planned / Parking Lot
20
25
 
21
- | Prio | Key | Title | Status | Current Spec | Notes |
22
- |---:|---|---|---|---|---|
23
- | P2 | | | `planned | review_needed` | | |
26
+ | Prio | Key | Title | Status | UR | PRD | SD | TP | QA | OR | Current Spec | Notes |
27
+ |---:|---|---|---|---|---|---|---|---|---|---|---|
28
+ | P2 | | | `planned | review_needed` | | | | | | | | |
24
29
 
25
30
  ## Completed / Superseded Pointers
26
31
 
@@ -0,0 +1,86 @@
1
+ # OR: <Title>
2
+
3
+ Gate: OR
4
+ Type: Orchestration Report
5
+ Report mode: `OR-lite | OR-full`
6
+ Status: `draft | done | superseded`
7
+
8
+ ## Run
9
+
10
+ - run_id:
11
+ - related_ur:
12
+ - related_prd:
13
+ - related_sd:
14
+ - related_tp:
15
+ - related_qa_report:
16
+ - mode_slice_decision:
17
+ - current_gate:
18
+ - decision: `pass | revise | block | in_progress`
19
+
20
+ ## Gate State
21
+
22
+ | Gate or step | Status | Evidence |
23
+ |---|---|---|
24
+ | UR | | |
25
+ | Brownfield Review | | |
26
+ | Mode/Slice Decision | | |
27
+ | PRD | | |
28
+ | SD | | |
29
+ | TP | | |
30
+ | Brownfield Analysis | | |
31
+ | CD+Tests | | |
32
+ | CR | | |
33
+ | QA | | |
34
+ | UAT | | |
35
+
36
+ ## Delivered
37
+
38
+ | Item | Evidence |
39
+ |---|---|
40
+ | | |
41
+
42
+ ## Not Delivered / Intentionally Deferred
43
+
44
+ | Item | Reason | Next owner or gate |
45
+ |---|---|---|
46
+ | | | |
47
+
48
+ ## Evidence
49
+
50
+ | Evidence | Source | Covers | Strength |
51
+ |---|---|---|---|
52
+ | | | | `direct | indirect | weak` |
53
+
54
+ ## Missing Evidence
55
+
56
+ | Missing evidence | Impact | Required next step |
57
+ |---|---|---|
58
+ | | `warn | revise | block` | |
59
+
60
+ ## Risks And Open Items
61
+
62
+ | Risk or open item | Impact | Owner or mitigation |
63
+ |---|---|---|
64
+ | | `warn | revise | block` | |
65
+
66
+ ## Context Graph Impact
67
+
68
+ - context_graph_impact: `none | link_only | update_existing_node | new_node_required | sot_drift`
69
+ - context_graph_refs:
70
+ - context_graph_required_action: `none | link | update | create | resolve_drift`
71
+ - context_graph_gate_effect: `none | warning | revise | block`
72
+ - context_graph_evidence:
73
+
74
+ ## Next Permissible Step
75
+
76
+ - next_allowed_action:
77
+ - required_approval:
78
+ - forbidden_until_then:
79
+
80
+ ## Quality Outlook
81
+
82
+ - quality_outlook:
83
+
84
+ ## Approval
85
+
86
+ OR does not approve later gates. It records the next permissible step.
@@ -0,0 +1,42 @@
1
+ # PRD: <Title>
2
+
3
+ Status: draft
4
+ Gate: PRD
5
+ Gate approval: open
6
+ Based on: UR
7
+ Date:
8
+ Owner:
9
+
10
+ ## 1. Product Scope
11
+
12
+ What exactly must be delivered?
13
+
14
+ ## 2. Acceptance Criteria
15
+
16
+ Which observable outcomes must be true?
17
+
18
+ ## 3. Non-Goals
19
+
20
+ What remains out of scope?
21
+
22
+ ## 4. Users And Roles
23
+
24
+ Who is affected and who decides?
25
+
26
+ ## 5. Constraints
27
+
28
+ Which business, operational, legal, security or compatibility constraints apply?
29
+
30
+ ## 6. Evidence Requirements
31
+
32
+ Which evidence must later support QA?
33
+
34
+ ## 7. Risks And Open Questions
35
+
36
+ Which questions must SD, TP or Brownfield Analysis clarify later?
37
+
38
+ ## 8. Next Step
39
+
40
+ Review this PRD and approve only with:
41
+
42
+ `Approval: PRD`
@@ -0,0 +1,38 @@
1
+ # QA Report: <Title>
2
+
3
+ Status: draft
4
+ Gate: QA
5
+ Gate approval: open
6
+ Based on: TP, implementation evidence, task-plan review
7
+ Date:
8
+ Owner:
9
+
10
+ ## 1. QA Decision
11
+
12
+ Decision: `pass | revise | block`
13
+
14
+ ## 2. TP Coverage
15
+
16
+ Which TP tasks are fully done, partially done or not done?
17
+
18
+ ## 3. Evidence
19
+
20
+ Which tests, logs, diffs, reviews or observations support the decision?
21
+
22
+ ## 4. Missing Evidence
23
+
24
+ What remains unverified?
25
+
26
+ ## 5. Risks
27
+
28
+ Which risks remain and what is their impact?
29
+
30
+ ## 6. Required Next Step
31
+
32
+ What is the next permitted action?
33
+
34
+ ## 7. Gate Approval
35
+
36
+ Approve this QA decision only with:
37
+
38
+ `Approval: QA`
@@ -0,0 +1,42 @@
1
+ # SD: <Title>
2
+
3
+ Status: draft
4
+ Gate: SD
5
+ Gate approval: open
6
+ Based on: PRD
7
+ Date:
8
+ Owner:
9
+
10
+ ## 1. Solution Overview
11
+
12
+ How will the approved PRD be solved?
13
+
14
+ ## 2. Ownership And Source Of Truth
15
+
16
+ Which existing modules, documents, services or owners remain authoritative?
17
+
18
+ ## 3. Architecture Decisions
19
+
20
+ Which design decisions are binding for implementation?
21
+
22
+ ## 4. Integration Points
23
+
24
+ Which APIs, data flows, UI surfaces, jobs, queues or external systems are affected?
25
+
26
+ ## 5. Constraints And Compatibility
27
+
28
+ Which constraints must implementation preserve?
29
+
30
+ ## 6. Test And Evidence Strategy
31
+
32
+ Which evidence must TP and QA later collect?
33
+
34
+ ## 7. Risks And Open Questions
35
+
36
+ Which risks remain before task planning?
37
+
38
+ ## 8. Next Step
39
+
40
+ Review this solution design and approve only with:
41
+
42
+ `Approval: SD`
@@ -0,0 +1,36 @@
1
+ # TP: <Title>
2
+
3
+ Status: draft
4
+ Gate: TP
5
+ Gate approval: open
6
+ Based on: SD
7
+ Date:
8
+ Owner:
9
+
10
+ ## 1. Task List
11
+
12
+ | task_id | Task | Acceptance mapping | Evidence required |
13
+ |---|---|---|---|
14
+ | T1 | | | |
15
+
16
+ ## 2. Test Plan
17
+
18
+ Which automated, manual or inspection checks must run?
19
+
20
+ ## 3. Brownfield Scope
21
+
22
+ Which existing owners, modules or artefacts must be inspected before implementation?
23
+
24
+ ## 4. Out Of Scope
25
+
26
+ Which implementation work is explicitly not part of this TP?
27
+
28
+ ## 5. Risks And Blockers
29
+
30
+ Which risks must block, revise or warn later QA?
31
+
32
+ ## 6. Next Step
33
+
34
+ Review this task and test plan and approve only with:
35
+
36
+ `Approval: TP`
@@ -0,0 +1,41 @@
1
+ # UR: <Title>
2
+
3
+ Status: draft
4
+ Gate: UR
5
+ Gate approval: open
6
+ Date:
7
+ Owner:
8
+
9
+ ## 1. Problem
10
+
11
+ What user, business or operational problem should be solved?
12
+
13
+ ## 2. Goal
14
+
15
+ What outcome should become possible?
16
+
17
+ ## 3. Scope
18
+
19
+ What is included in this first slice?
20
+
21
+ ## 4. Non-Goals
22
+
23
+ What is explicitly not part of this UR?
24
+
25
+ ## 5. Acceptance Signals
26
+
27
+ How will we know the need is clear enough for PRD?
28
+
29
+ ## 6. Existing Source Of Truth
30
+
31
+ Which existing repository artefacts, docs, code paths or decisions must be respected?
32
+
33
+ ## 7. Risks And Unknowns
34
+
35
+ Which questions must Brownfield Review, PRD, SD or later implementation-preparation Brownfield Analysis clarify?
36
+
37
+ ## 8. Next Step
38
+
39
+ Review this UR and approve only with:
40
+
41
+ `Approval: UR`
@@ -30,17 +30,26 @@ Use Quick Task Mode for questions, small reviews, local debugging and narrow fix
30
30
 
31
31
  Use Structured Delivery Mode for new capabilities, architecture, policy, persistence, release-critical work, formal artefacts or explicit approvals.
32
32
 
33
+ Any new product semantics, functional change or user-visible behaviour change requires a durable UR first.
34
+ Persist it in `.agdf/control/` or link it from there to the target repository's authoritative source of truth before PRD, SD, TP, Brownfield Analysis or implementation.
35
+ Treat approval text and durable artefact presence as separate checks for UR, PRD, SD, TP and QA report decisions.
36
+ Do not infer approval from "ok", "go ahead", "do it", "approved", "continue", "leg los" or similar wording.
37
+ Those phrases may express intent to proceed, but they do not unlock a gate unless the exact gate formula is present.
38
+ After `Approval: UR`, run lightweight Brownfield Review before PRD when Brownfield, ownership, runtime, policy, persistence, architecture, UI or UX impact is possible.
39
+ Brownfield Review must produce a Mode/Slice Decision: `quick_task`, `structured_slice`, `structured_delivery` or `block`.
40
+ Do not assume the full PRD/SD/TP chain before that decision; use only as much gate depth as the reviewed change size justifies.
41
+
33
42
  Quick Task Mode must still cite evidence and close with a concrete result.
34
43
  Structured Delivery Mode must respect gates, reviews and closeout discipline.
35
44
 
36
45
  ## Skill Routing
37
46
  | Skill | Use For | Boundary |
38
47
  |---|---|---|
39
- | `brownfield-analysis` | after gate-check permits implementation preparation, before non-trivial changes in existing systems | clarifies reuse, owners and risks; never bypasses gate-check |
48
+ | `brownfield-analysis` | after gate-check permits Brownfield Review or implementation preparation, before non-trivial changes in existing systems | clarifies reuse, owners, risks and Mode/Slice Decision; never bypasses gate-check; Brownfield Review is not implementation permission |
40
49
  | `clean-implementation-review` | inspect workarounds, fallbacks, parallel structures or symptom fixes | not a TP or QA substitute |
41
50
  | `code-review` | mandatory CR step after code changes, focused on defects, regression and security findings | does not replace TP review, clean review or QA |
42
51
  | `delivery-closeout` | commit/PR-near handoff after QA/OR/UAT | never performs VCS actions automatically |
43
- | `gate-check` | new build/change intent, unclear approval, Structured Delivery, later-gate artefact requested | does not create later artefacts |
52
+ | `gate-check` | new build/change intent, unclear approval, Structured Delivery, later-gate artefact requested | does not create later artefacts or skip Mode/Slice Decision after Brownfield Review |
44
53
  | `qa-gate` | final QA decision | only instance for `pass | revise | block` |
45
54
  | `release-or` | auditable closeout for every relevant run | not a QA substitute |
46
55
  | `task-plan-review` | after implementation and before QA, verify TP coverage | no final QA decision |
@@ -48,6 +57,7 @@ Structured Delivery Mode must respect gates, reviews and closeout discipline.
48
57
  Select exactly one primary skill first.
49
58
  Add more only when they cover a distinct concrete risk dimension.
50
59
  Do not choose `brownfield-analysis` as the first primary skill for a fresh "I want to build/change X" prompt unless `gate-check` or existing live AGDF control state already makes implementation preparation the next allowed action.
60
+ Never jump directly from `Approval: UR`, implicit consent, or a generic "start" request to implementation. Route to Brownfield Review, then Mode/Slice Decision, then the smallest safe next gate or Quick Task execution.
51
61
 
52
62
  ## Runtime Contract
53
63
  For repeated output, gate, Quality Contract and Context Graph rules, use:
@@ -62,6 +72,9 @@ Use templates as starting points only.
62
72
 
63
73
  Do not let chat history become the source of truth for gate state, approvals, evidence, backlog status or durable Brownfield knowledge.
64
74
 
75
+ AGDF is a native agent workflow first: read the live control state, apply the active skill and make the next allowed step explicit.
76
+ Use `doctor`, `gate-check --json` or `delivery-map --json` as validators when a machine-readable proof is useful, not as a substitute for the router or skills.
77
+
65
78
  ## Closeout
66
79
  After relevant code, documentation, skill or governance changes:
67
80