bmad-handoff

Name: bmad-handoff
Author: aj-geddes

467stars73forks

How to install bmad-handoff?

aj-geddes/claude-code-bmad-skills/bmad-handoff

$npx -y skills add aj-geddes/claude-code-bmad-skills --skill bmad-handoff

Installs into the current project.

Summary

Emits a dev-tool-agnostic handoff manifest from ready-for-dev stories so an external dev plugin or runner can pick up and execute the work. Use when the user says "generate a handoff", "create handoff manifest", "export stories for dev", "hand off to dev tool", "produce handoff m

Files · 1

View on GitHub

SKILL.md

1	# BMAD Handoff
2
3	Purpose: Scan the planning output folder for all stories at status `ready-for-dev`,
4	compile them into a single `handoff-manifest.json`, and leave it where any downstream
5	dev tool can read it — without coupling to any specific runner.
6
7	This is the last artifact the planning plugin produces. What happens next is owned by
8	the external dev tool, not by this skill.
9
10	## When to Run
11
12	Run this skill after the story-writing phase is complete and you (or the user) have
13	confirmed that at least one story carries status `ready-for-dev`. The manifest is a
14	point-in-time snapshot; re-run the skill to refresh it.
15
16	## Workflow
17
18	Use TodoWrite to track progress through these steps.
19
20	### 1. Locate the output folder
21
22	Look for `bmad-output/project-context.md` (the default output folder) or ask the
23	user for the output folder. Default: `bmad-output/`.
24
25	### 2. Discover story files
26
27	Glob for `*/{epic}.{story}..story.md` under the output folder.
28	Accept alternative flat layouts (`stories/*.story.md`) if the glob turns up nothing.
29
30	### 3. Filter to ready-for-dev
31
32	Read the `Status:` header field of each story file (in the story header block, not a
33	`## Status` heading). Include only stories whose status value is exactly `ready-for-dev`.
34
35	If none are found, report which statuses were seen and stop — do not produce an
36	empty manifest.
37
38	### 4. Extract per-story fields
39
40	For each qualifying story file, extract:
41
42	\| Field \| Source in story file \|
43	\|---\|---\|
44	\| `id` \| Filename stem or `## Story` heading ID \|
45	\| `storyFilePath` \| Relative path from project root \|
46	\| `status` \| `Status:` header field value \|
47	\| `epic` \| First segment of filename, e.g. `"2"` in `2.1.stripe.story.md` \|
48	\| `storyNumber` \| Second segment, e.g. `"1"` \|
49	\| `title` \| First H1 or `## Story` heading text \|
50	\| `ownedScope` \| `## Owned File/Module Scope` — list every path verbatim \|
51	\| `wave` \| `## Dependency Maps` → wave/parallel_set annotation (integer or null) \|
52	\| `parallelSet` \| Same section — parallel set label if present (string or null) \|
53	\| `dependencies` \| `## Dependency Maps` → blocked-by story IDs (array, may be empty) \|
54	\| `acceptanceCriteriaSummary` \| First 3 AC items from `## Acceptance Criteria`, each ≤120 chars \|
55	\| `lockedSectionsNote` \| Static string — see schema \|
56	\| `devAgentRecord` \| Static null — placeholder for the dev tool to populate \|
57
58	### 5. Compute wave order
59
60	If stories do not already carry explicit wave annotations:
61	- Stories with empty `dependencies` arrays are wave 1.
62	- A story whose every dependency is in wave N or lower is wave N+1.
63	- Add the computed `wave` value; leave `parallelSet` null when not annotated.
64
65	### 6. Write the manifest
66
67	Write to `{outputFolder}/handoff-manifest.json`.
68
69	Use the schema from
70	`${CLAUDE_PLUGIN_ROOT}/skills/bmad-handoff/templates/handoff-manifest.schema.json`
71	as the structural contract. Populate `schemaVersion: "1.0"`.
72
73	Sort stories by `wave` ascending, then by `id` ascending within each wave.
74
75	### 7. Report to the user
76
77	Print a compact summary:
78
79	```
80	Handoff manifest written → bmad-output/handoff-manifest.json
81	schemaVersion : 1.0
82	stories : <N> ready-for-dev
83	waves : <W> (wave 1 has <X> stories, can start immediately)
84	output path : bmad-output/handoff-manifest.json
85	```
86
87	If any story was missing required sections (e.g. no `## Owned File/Module Scope`),
88	list those as warnings — do not silently omit or fabricate data.
89
90	## Manifest Field Definitions (Quick Reference)
91
92	See REFERENCE.md for the full schema narrative and adapter notes.
93
94	\| Field \| Type \| Required \| Notes \|
95	\|---\|---\|---\|---\|
96	\| `schemaVersion` \| string \| yes \| Semver string; current = `"1.0"` \|
97	\| `generatedAt` \| string \| yes \| ISO-8601 UTC timestamp \|
98	\| `projectName` \| string \| yes \| From project-context.md or user input \|
99	\| `outputFolder` \| string \| yes \| Relative path used to find stories \|
100	\| `stories` \| array \| yes \| One object per ready-for-dev story \|
101	\| `stories[].id` \| stri

People who liked this also viewed

6 picks

Type

bmad-handoff

467stars73forks

View on GitHub

How to install bmad-handoff?

aj-geddes/claude-code-bmad-skills/bmad-handoff

$npx -y skills add aj-geddes/claude-code-bmad-skills --skill bmad-handoff

Installs into the current project.

Summary

Files · 1

View on GitHub

SKILL.md

1	# BMAD Handoff
2
3	Purpose: Scan the planning output folder for all stories at status `ready-for-dev`,
4	compile them into a single `handoff-manifest.json`, and leave it where any downstream
5	dev tool can read it — without coupling to any specific runner.
6
7	This is the last artifact the planning plugin produces. What happens next is owned by
8	the external dev tool, not by this skill.
9
10	## When to Run
11
12	Run this skill after the story-writing phase is complete and you (or the user) have
13	confirmed that at least one story carries status `ready-for-dev`. The manifest is a
14	point-in-time snapshot; re-run the skill to refresh it.
15
16	## Workflow
17
18	Use TodoWrite to track progress through these steps.
19
20	### 1. Locate the output folder
21
22	Look for `bmad-output/project-context.md` (the default output folder) or ask the
23	user for the output folder. Default: `bmad-output/`.
24
25	### 2. Discover story files
26
27	Glob for `*/{epic}.{story}..story.md` under the output folder.
28	Accept alternative flat layouts (`stories/*.story.md`) if the glob turns up nothing.
29
30	### 3. Filter to ready-for-dev
31
32	Read the `Status:` header field of each story file (in the story header block, not a
33	`## Status` heading). Include only stories whose status value is exactly `ready-for-dev`.
34
35	If none are found, report which statuses were seen and stop — do not produce an
36	empty manifest.
37
38	### 4. Extract per-story fields
39
40	For each qualifying story file, extract:
41
42	\| Field \| Source in story file \|
43	\|---\|---\|
44	\| `id` \| Filename stem or `## Story` heading ID \|
45	\| `storyFilePath` \| Relative path from project root \|
46	\| `status` \| `Status:` header field value \|
47	\| `epic` \| First segment of filename, e.g. `"2"` in `2.1.stripe.story.md` \|
48	\| `storyNumber` \| Second segment, e.g. `"1"` \|
49	\| `title` \| First H1 or `## Story` heading text \|
50	\| `ownedScope` \| `## Owned File/Module Scope` — list every path verbatim \|
51	\| `wave` \| `## Dependency Maps` → wave/parallel_set annotation (integer or null) \|
52	\| `parallelSet` \| Same section — parallel set label if present (string or null) \|
53	\| `dependencies` \| `## Dependency Maps` → blocked-by story IDs (array, may be empty) \|
54	\| `acceptanceCriteriaSummary` \| First 3 AC items from `## Acceptance Criteria`, each ≤120 chars \|
55	\| `lockedSectionsNote` \| Static string — see schema \|
56	\| `devAgentRecord` \| Static null — placeholder for the dev tool to populate \|
57
58	### 5. Compute wave order
59
60	If stories do not already carry explicit wave annotations:
61	- Stories with empty `dependencies` arrays are wave 1.
62	- A story whose every dependency is in wave N or lower is wave N+1.
63	- Add the computed `wave` value; leave `parallelSet` null when not annotated.
64
65	### 6. Write the manifest
66
67	Write to `{outputFolder}/handoff-manifest.json`.
68
69	Use the schema from
70	`${CLAUDE_PLUGIN_ROOT}/skills/bmad-handoff/templates/handoff-manifest.schema.json`
71	as the structural contract. Populate `schemaVersion: "1.0"`.
72
73	Sort stories by `wave` ascending, then by `id` ascending within each wave.
74
75	### 7. Report to the user
76
77	Print a compact summary:
78
79	```
80	Handoff manifest written → bmad-output/handoff-manifest.json
81	schemaVersion : 1.0
82	stories : <N> ready-for-dev
83	waves : <W> (wave 1 has <X> stories, can start immediately)
84	output path : bmad-output/handoff-manifest.json
85	```
86
87	If any story was missing required sections (e.g. no `## Owned File/Module Scope`),
88	list those as warnings — do not silently omit or fabricate data.
89
90	## Manifest Field Definitions (Quick Reference)
91
92	See REFERENCE.md for the full schema narrative and adapter notes.
93
94	\| Field \| Type \| Required \| Notes \|
95	\|---\|---\|---\|---\|
96	\| `schemaVersion` \| string \| yes \| Semver string; current = `"1.0"` \|
97	\| `generatedAt` \| string \| yes \| ISO-8601 UTC timestamp \|
98	\| `projectName` \| string \| yes \| From project-context.md or user input \|
99	\| `outputFolder` \| string \| yes \| Relative path used to find stories \|
100	\| `stories` \| array \| yes \| One object per ready-for-dev story \|
101	\| `stories[].id` \| stri

People who liked this also viewed

6 picks

Type