flow-workflows 0.81.0 → 0.82.0
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.
- package/adapters/codex/CORE.md +120 -6
- package/adapters/codex/skills/flow-bug-fix/SKILL.md +8 -2
- package/adapters/codex/skills/flow-bug-investigate/SKILL.md +5 -0
- package/adapters/codex/skills/flow-bug-postmortem/SKILL.md +1 -1
- package/adapters/codex/skills/flow-bug-review/SKILL.md +1 -1
- package/adapters/codex/skills/flow-bug-ship/SKILL.md +15 -2
- package/adapters/codex/skills/flow-bug-start/SKILL.md +9 -0
- package/adapters/codex/skills/flow-bug-validate/SKILL.md +2 -2
- package/adapters/codex/skills/flow-doctor/SKILL.md +11 -3
- package/adapters/codex/skills/flow-feat-build/SKILL.md +12 -6
- package/adapters/codex/skills/flow-feat-design/SKILL.md +6 -0
- package/adapters/codex/skills/flow-feat-review/SKILL.md +4 -1
- package/adapters/codex/skills/flow-feat-ship/SKILL.md +41 -3
- package/adapters/codex/skills/flow-feat-start/SKILL.md +14 -0
- package/adapters/codex/skills/flow-feat-validate/SKILL.md +12 -4
- package/adapters/codex/skills/flow-init/SKILL.md +4 -3
- package/adapters/codex/skills/flow-work-query/SKILL.md +1 -1
- package/adapters/codex/skills/flow-work-resume/SKILL.md +7 -1
- package/adapters/gemini/CORE.md +120 -6
- package/adapters/gemini/commands/flow/bug/fix.toml +8 -2
- package/adapters/gemini/commands/flow/bug/investigate.toml +5 -0
- package/adapters/gemini/commands/flow/bug/postmortem.toml +1 -1
- package/adapters/gemini/commands/flow/bug/review.toml +1 -1
- package/adapters/gemini/commands/flow/bug/ship.toml +15 -2
- package/adapters/gemini/commands/flow/bug/start.toml +9 -0
- package/adapters/gemini/commands/flow/bug/validate.toml +2 -2
- package/adapters/gemini/commands/flow/doctor.toml +11 -3
- package/adapters/gemini/commands/flow/feat/build.toml +12 -6
- package/adapters/gemini/commands/flow/feat/design.toml +6 -0
- package/adapters/gemini/commands/flow/feat/review.toml +4 -1
- package/adapters/gemini/commands/flow/feat/ship.toml +41 -3
- package/adapters/gemini/commands/flow/feat/start.toml +14 -0
- package/adapters/gemini/commands/flow/feat/validate.toml +12 -4
- package/adapters/gemini/commands/flow/init.toml +4 -3
- package/adapters/gemini/commands/flow/work/query.toml +1 -1
- package/adapters/gemini/commands/flow/work/resume.toml +7 -1
- package/adapters/hermes/CORE.md +120 -6
- package/adapters/hermes/skills/flow-bug-fix/SKILL.md +8 -2
- package/adapters/hermes/skills/flow-bug-investigate/SKILL.md +5 -0
- package/adapters/hermes/skills/flow-bug-postmortem/SKILL.md +1 -1
- package/adapters/hermes/skills/flow-bug-review/SKILL.md +1 -1
- package/adapters/hermes/skills/flow-bug-ship/SKILL.md +15 -2
- package/adapters/hermes/skills/flow-bug-start/SKILL.md +9 -0
- package/adapters/hermes/skills/flow-bug-validate/SKILL.md +2 -2
- package/adapters/hermes/skills/flow-doctor/SKILL.md +11 -3
- package/adapters/hermes/skills/flow-feat-build/SKILL.md +12 -6
- package/adapters/hermes/skills/flow-feat-design/SKILL.md +6 -0
- package/adapters/hermes/skills/flow-feat-review/SKILL.md +4 -1
- package/adapters/hermes/skills/flow-feat-ship/SKILL.md +41 -3
- package/adapters/hermes/skills/flow-feat-start/SKILL.md +14 -0
- package/adapters/hermes/skills/flow-feat-validate/SKILL.md +12 -4
- package/adapters/hermes/skills/flow-init/SKILL.md +4 -3
- package/adapters/hermes/skills/flow-work-query/SKILL.md +1 -1
- package/adapters/hermes/skills/flow-work-resume/SKILL.md +7 -1
- package/adapters/opencode/CORE.md +120 -6
- package/adapters/opencode/commands/flow-bug-fix.md +8 -2
- package/adapters/opencode/commands/flow-bug-investigate.md +5 -0
- package/adapters/opencode/commands/flow-bug-postmortem.md +1 -1
- package/adapters/opencode/commands/flow-bug-review.md +1 -1
- package/adapters/opencode/commands/flow-bug-ship.md +15 -2
- package/adapters/opencode/commands/flow-bug-start.md +9 -0
- package/adapters/opencode/commands/flow-bug-validate.md +2 -2
- package/adapters/opencode/commands/flow-doctor.md +11 -3
- package/adapters/opencode/commands/flow-feat-build.md +12 -6
- package/adapters/opencode/commands/flow-feat-design.md +6 -0
- package/adapters/opencode/commands/flow-feat-review.md +4 -1
- package/adapters/opencode/commands/flow-feat-ship.md +41 -3
- package/adapters/opencode/commands/flow-feat-start.md +14 -0
- package/adapters/opencode/commands/flow-feat-validate.md +12 -4
- package/adapters/opencode/commands/flow-init.md +4 -3
- package/adapters/opencode/commands/flow-work-query.md +1 -1
- package/adapters/opencode/commands/flow-work-resume.md +7 -1
- package/package.json +1 -1
- package/plugins/flow/.claude-plugin/plugin.json +1 -1
- package/plugins/flow/CHANGELOG.md +22 -0
- package/plugins/flow/examples/FLOW.template.md +12 -3
|
@@ -22,6 +22,37 @@ Read `~/.claude/flow/CORE.opencode.md` first (shared rules: `FLOW.md` step 0, mo
|
|
|
22
22
|
|
|
23
23
|
Closes the feature: commit, push, MR/PR (assigned per `git.assignee`, squash per `git.squash`, sections per `git.request_sections`) and an optional offer to consolidate knowledge.
|
|
24
24
|
|
|
25
|
+
**In `unattended`** (flow-core §2.1) this command asks nothing, and it is the only place that run
|
|
26
|
+
pushes. Each place it would ask:
|
|
27
|
+
|
|
28
|
+
- **The MR/PR itself (§3–§4)** — `autonomy.unattended_ship: draft` → skip the confirmation and
|
|
29
|
+
create it **as draft** with the §2 content (§4.1 says how); never merge it, never mark it ready.
|
|
30
|
+
The body gains one line saying an unattended run opened it and nobody confirmed its brief.
|
|
31
|
+
Empty or `stop` → nothing is pushed: question `ship`, the §3 preview as the question, and the run
|
|
32
|
+
ends; a person ships it from a session of their own. Pre-deploy SQL in the branch → question
|
|
33
|
+
`migration`, whatever `unattended_ship` says.
|
|
34
|
+
- **The §1 tree check** — `reviewed_sha` or `validated_sha` behind `HEAD` with more than tests →
|
|
35
|
+
re-run what is behind, **once per MR/PR**: the review on the delta, then `validate`. First record
|
|
36
|
+
`{ "key": "unattended:ship-recheck", "default": "mr <n>", "phase": "ship" }` in
|
|
37
|
+
`meta.json.defaults_used[]` — the record, not the context, is what remembers it; an entry with this
|
|
38
|
+
MR/PR's `n` already there → `blocked`, whatever the sha. Tracked changes this run's own phases left uncommitted → commit them
|
|
39
|
+
as WIP first (a WIP commit is `auto`'s to make) and check again · **blocking TODO/FIXME (§1)** →
|
|
40
|
+
continue and list them in the body.
|
|
41
|
+
- **Posting anything else** — the performance comment (§4.3), a contract handoff (§6.3) — → not
|
|
42
|
+
posted; the numbers go in the body, the handoff stays `pending`.
|
|
43
|
+
- **The offers and surveys (§5, §5.1, §5.2, §5.5, §6.4, §6.4bis)** → `Later`; nothing written outside
|
|
44
|
+
the work folder.
|
|
45
|
+
- **"Was it merged?" (§6.1)** → not merged; `meta.json` as §6.1 says for an unmerged MR/PR. **The
|
|
46
|
+
parent merged first (§6.2.1)** → blocked, never force-push. **Train continuation (§6.2)** → not
|
|
47
|
+
started: one MR/PR per unattended run.
|
|
48
|
+
- **At Close** write the stop file with `reason: "done"`, `mr_url`, and — with `mrs` still pending —
|
|
49
|
+
`detail` naming the next MR/PR and what it waits on (its `depends_on` merged, then
|
|
50
|
+
`/flow-feat-build`). It is the last thing the run does.
|
|
51
|
+
|
|
52
|
+
**In every mode**, criteria `07-validation.md` marks `not-verified-unattended` go into the body under
|
|
53
|
+
`## Not verified — needs a human`, with what to do for each. The §3 preview shows that section like
|
|
54
|
+
any other, and the person reading it is the check the run could not do.
|
|
55
|
+
|
|
25
56
|
## 1. Pre-flight
|
|
26
57
|
|
|
27
58
|
- Read `meta.json` and `00-summary.md`; open in full only the `Brief MR/PR #N` of `05-implementation.md`, the `## Summary`/blockers of `06-review.md` and `07-validation.md`. (flow-core §5)
|
|
@@ -133,7 +164,7 @@ Determine whether the branch modifies the database (migrations, mappings/schema,
|
|
|
133
164
|
|
|
134
165
|
## 3. Show the user and wait for confirmation (MANDATORY)
|
|
135
166
|
|
|
136
|
-
**Never skip this step, even when the content seems obvious.** Hard gate in every mode.
|
|
167
|
+
**Never skip this step, even when the content seems obvious.** Hard gate in every mode — in `unattended`, the unattended paragraph above decides it.
|
|
137
168
|
|
|
138
169
|
Print to the user in this exact format:
|
|
139
170
|
|
|
@@ -174,7 +205,14 @@ git rev-parse --abbrev-ref --symbolic-full-name @{u} 2>/dev/null # must NOT be
|
|
|
174
205
|
- Train mode (`stacked_on` ≠ null) → the MR/PR must target that parent branch, not the main base.
|
|
175
206
|
|
|
176
207
|
### 4.1 Create MR/PR
|
|
177
|
-
|
|
208
|
+
**A draft (`unattended`, flow-core §2.1) — read this first: it replaces the invocation below.** Do
|
|
209
|
+
**not** invoke `commit-push-pr`, and do not use the plain CLI fallback below: both open a normal
|
|
210
|
+
MR/PR. Commit, `git push -u origin HEAD`, and create it with the forge CLI and its draft flag
|
|
211
|
+
(`gh pr create --draft`, `glab mr create --draft`, the equivalent elsewhere), assignee and squash as
|
|
212
|
+
below. The forge refusing a draft → `blocked` with its message; never retried without the flag.
|
|
213
|
+
Then record the URL, as the paragraph after the bullets says.
|
|
214
|
+
|
|
215
|
+
Otherwise, only here — with the content approved in §3 — invoke `Skill commit-commands:commit-push-pr` passing **the final title and description**. The skill must not re-ask for the content; if it does, answer with what was confirmed. Any push it does must be `git push -u origin HEAD` (own branch), never to the base branch.
|
|
178
216
|
|
|
179
217
|
- `git.assignee` not empty → assign to that user. `git.squash` `true` → enable squash-before-merge.
|
|
180
218
|
- `commit-push-pr` skill unavailable → commit and push manually and create the MR/PR with the `git.cli` CLI from `FLOW.md` — always with the content confirmed in §3.
|
|
@@ -467,7 +505,7 @@ Read `meta.json.conventions_candidates[]` (flow-core §8). **No entries with `st
|
|
|
467
505
|
|
|
468
506
|
### 6.4ter Defaults that were used (one line, no question)
|
|
469
507
|
|
|
470
|
-
`meta.json.defaults_used[]` non-empty → **one line** in the Close: how many empty `FLOW.md` keys this
|
|
508
|
+
`meta.json.defaults_used[]` non-empty (keys prefixed `unattended:` not counted) → **one line** in the Close: how many empty `FLOW.md` keys this
|
|
471
509
|
work resolved with their default, and that `/flow-doctor` lists which ones and prints the line to
|
|
472
510
|
paste. Never a list here, and never a question: in `guided`/`auto` not asking is the contract
|
|
473
511
|
(flow-core §0, §2), and in `manual` the phases already offered what they were allowed to offer.
|
|
@@ -25,6 +25,20 @@ Read `~/.claude/flow/CORE.opencode.md` first (shared rules: `FLOW.md` step 0, mo
|
|
|
25
25
|
- **Given** — an identifier in `tracker.prefix` format from `FLOW.md` → *ticket mode* (§2 reads it).
|
|
26
26
|
- **Empty** — *ticket-less mode*: do **not** stop or demand a ticket. Draft a ticket from the conversation just held (§2.5), as `/flow-feat-ship` builds the MR/PR body from the work log. Ask for a one-liner only if there is no conversation to draft from.
|
|
27
27
|
|
|
28
|
+
**In `unattended`** (flow-core §2.1) this command asks nothing; each place it would ask resolves as
|
|
29
|
+
that table says. Here: a ticket is required — empty, or a tracker read that fails, ends the run
|
|
30
|
+
`blocked` (there is no conversation to draft from and nobody to paste it) · uncommitted changes or a
|
|
31
|
+
branch other than the main one (§5.1) → `blocked`: the runner owns the checkout · cross-repo scope
|
|
32
|
+
(§3.5) → only the repos the ticket names · below-XS (§4) → open the work · `git.worktree: ask`
|
|
33
|
+
(§5.0) → in place · §5.5's linked branch → skipped, it pushes a branch and nothing is pushed before
|
|
34
|
+
`ship` · the tracker transition (§6.5) → as `auto`. **Clarifying questions (§2.1 precedence, §3)
|
|
35
|
+
wait until the branch and the work folder exist** — create them (§5, §6) first, write the questions
|
|
36
|
+
into `01-context.md` under "Decisions clarified at start" as open, then question `clarify` with all
|
|
37
|
+
of them in one stop. Whoever picks the work up is asked them by the next phase, first (flow-core
|
|
38
|
+
§2.1).
|
|
39
|
+
A `blocked` before the folder exists still has the stop file: its path does not depend on the folder
|
|
40
|
+
(flow-core §2.1).
|
|
41
|
+
|
|
28
42
|
## 1. Pre-flight
|
|
29
43
|
|
|
30
44
|
- Read the effective FLOW configuration per flow-core §0; neither base nor active overlay → each
|
|
@@ -151,12 +151,17 @@ Verify `needs-manual` criteria **together with the user**, one or a few at a tim
|
|
|
151
151
|
- **Blocked** → status `unproven`; record why. Does not pass the gate.
|
|
152
152
|
4. Repeat until every `needs-manual` criterion is `proven-manually`, or the user decides to stop (the rest stay `unproven`, which blocks advancing in §7).
|
|
153
153
|
|
|
154
|
+
**In `unattended`** there is nobody to run the check (flow-core §2.1): each `needs-manual` criterion
|
|
155
|
+
gets status `not-verified-unattended` with what a person must do to verify it, and that status does
|
|
156
|
+
**not** block §7 — the MR/PR can only be a draft, and `ship` lists these criteria in its body under
|
|
157
|
+
"Not verified — needs a human". A criterion §3.5 could have driven and did not is still `unproven`.
|
|
158
|
+
|
|
154
159
|
## 5. Manual edge cases
|
|
155
160
|
|
|
156
161
|
If the feature has UI or critical flows:
|
|
157
162
|
- Payments → test with the test cards or credentials the provider publishes for its sandbox (and the harness skill for that provider, if installed).
|
|
158
163
|
- Workers/queues → make sure no jobs are stuck in dead-letter. Stuck jobs that are not yours: do not touch them here.
|
|
159
|
-
- Migrations → run `quality.db_update` from `FLOW.md` (if defined); verify no unexpected schema difference with the comparison command the project uses.
|
|
164
|
+
- Migrations → run `quality.db_update` from `FLOW.md` (if defined; in `unattended`, never — a migration to apply is question `migration`, flow-core §2.1); verify no unexpected schema difference with the comparison command the project uses.
|
|
160
165
|
|
|
161
166
|
## 6. Output
|
|
162
167
|
|
|
@@ -210,11 +215,14 @@ Write `.claude/work/<TICKET>/07-validation.md`:
|
|
|
210
215
|
## 7. Close
|
|
211
216
|
|
|
212
217
|
- **Do not advance `phase`** if tests are red, regressions are found, or **any acceptance criterion is `unproven`** (no test demonstrably asserts it, and neither §3.5 nor the user confirmed it). The criterion→test mapping is part of the gate — the same "do not advance on red" rule. The user resolves the gap (add the missing test, fix the implementation, or finish the manual verification) and returns to `/flow-feat-validate`.
|
|
213
|
-
- Suite green **and** every acceptance criterion `proven-by-test`, `proven-by-agent` or `proven-manually` → `phase = "validate"`, add to `phases_done`. **Multi-MR/PR work** → also add `validate` to the current `in_progress` MR/PR's own `phases_done` (its `mrs[]` entry) — the per-MR/PR marker `/flow-feat-ship §1` gates on. Suggest `/flow-feat-ship`.
|
|
218
|
+
- Suite green **and** every acceptance criterion `proven-by-test`, `proven-by-agent` or `proven-manually` (in `unattended`, also `not-verified-unattended`, §4) → `phase = "validate"`, add to `phases_done`. **Multi-MR/PR work** → also add `validate` to the current `in_progress` MR/PR's own `phases_done` (its `mrs[]` entry) — the per-MR/PR marker `/flow-feat-ship §1` gates on. Suggest `/flow-feat-ship`.
|
|
214
219
|
- **Record *what* you validated**, in the same write and only when the phase advances: `validated_sha` = `git rev-parse HEAD`, work-level and in this MR/PR's `mrs[]` entry. `/flow-feat-ship §1` reads it to tell "the suite passed here" from "the suite passed on something else".
|
|
215
220
|
- **Unchecked edge cases and open risks become `followups[]` entries** (flow-core §7) before advancing — see the output template above.
|
|
216
221
|
- **Stage what the validation taught** (`knowledge.stage` set; silence by default): a mutation that stayed green, a disguise the scenario wore, a fixture that could not produce the case — findings about how this codebase is tested, not about this diff. Same rule and evidence line as the review's Close.
|
|
217
222
|
- Overwrite `00-summary.md` whole (≤15 lines, flow-core §5).
|
|
218
|
-
- **Autonomy handoff — this one stops in every mode, `auto` included.** `ship` pushes and opens the MR/PR: a hard gate in every mode, so do **not** chain into it. Stop here and propose `/flow-feat-ship` with a single `AskUserQuestion` (recommended option by default), invoking it only when the user confirms. This is the deliberate end of
|
|
223
|
+
- **Autonomy handoff — this one stops in every mode, `auto` included.** `ship` pushes and opens the MR/PR: a hard gate in every mode, so do **not** chain into it. Stop here and propose `/flow-feat-ship` with a single `AskUserQuestion` (recommended option by default), invoking it only when the user confirms. This is the deliberate end of a run in `auto`.
|
|
224
|
+
**In `unattended`** (flow-core §2.1): the phase advanced → chain into `/flow-feat-ship`, which
|
|
225
|
+
decides from `autonomy.unattended_ship` between a draft MR/PR and a question. It did not advance
|
|
226
|
+
(red suite, a criterion `unproven`) → `blocked`, naming what failed; never chain.
|
|
219
227
|
|
|
220
|
-
In `guided`/`auto` this is, together with the brief in `/flow-feat-build §2`, **the only stop of the whole MR/PR** — everything between them ran
|
|
228
|
+
In `guided`/`auto` this is, together with the brief in `/flow-feat-build §2`, **the only stop of the whole MR/PR** — everything between them ran without a person. It carries the **full stop header** and the body answers, in this order: what is green, what this MR/PR proves and what it does not, and what shipping it takes. Assume the user has read none of the build and review.
|
|
@@ -96,9 +96,10 @@ review, not a questionnaire. Then, and only then, five questions:
|
|
|
96
96
|
in the confirmation screen, not a question.
|
|
97
97
|
3. **Autonomy** (`autonomy.mode`): `manual` *(Recommended)* — stops at each decision, only recommends
|
|
98
98
|
the next command; `guided` — resolves low-risk decisions and chains phases, asks at the real ones;
|
|
99
|
-
`auto` — also resolves the rest with recorded defaults. Explain the hard gates that stop in
|
|
100
|
-
|
|
101
|
-
`auto`
|
|
99
|
+
`auto` — also resolves the rest with recorded defaults. Explain the hard gates that stop in these
|
|
100
|
+
three modes and the list of what `guided`/`auto` never ask (both in the `~/.claude/flow/CORE.opencode.md` §2),
|
|
101
|
+
so `auto` runs without a stop between the brief and `ship`. Do not offer `unattended`: it is for a
|
|
102
|
+
runner, written by hand where only the runner reads it. Editable any time. Empty = `manual`.
|
|
102
103
|
4. **The quality commands as one block** (`quality.test`, `test_one`, `static_analysis`, `style_fix`,
|
|
103
104
|
`db_update`, `db_diff`) — confirm or edit what §2 found, all in one question, each may stay empty
|
|
104
105
|
(= auto-discover at the moment it is needed). Several suites chain into `test` with `&&`.
|
|
@@ -24,7 +24,7 @@ Read `~/.claude/flow/CORE.opencode.md` first (shared rules: `FLOW.md` step 0, mo
|
|
|
24
24
|
|
|
25
25
|
Invoked by `/flow-feat-review` and `/flow-bug-review` (duel over the diff), `/flow-work-respond` (a reviewer objected to a query), `/flow-feat-design` (access paths), `/flow-feat-build`, `/flow-feat-validate`. Invoke directly for any doubt outside them (chat question, comment on an open MR/PR, a year-old query).
|
|
26
26
|
|
|
27
|
-
**Autonomy.** Modes as in flow-core §2. The duel, challenger count and whether to measure are flow mechanics — never a question, in any mode. **Hard gates — stop and ask, in every mode:** (1) **creating, seeding or dropping any database, schema or table**, including a throwaway one (§4) — show exact commands and target, never a database the project uses; (2) **any DDL** on a real database (index, column, collation) — the flow's schema gate; (3) editing code (output is a verdict, not a patch); (4) posting anything to an MR/PR or ticket.
|
|
27
|
+
**Autonomy.** Modes as in flow-core §2. The duel, challenger count and whether to measure are flow mechanics — never a question, in any mode. **Hard gates — stop and ask, in every mode:** (1) **creating, seeding or dropping any database, schema or table**, including a throwaway one (§4) — show exact commands and target, never a database the project uses; (2) **any DDL** on a real database (index, column, collation) — the flow's schema gate; (3) editing code (output is a verdict, not a patch); (4) posting anything to an MR/PR or ticket. **In `unattended`** (flow-core §2.1) (1) and (2) are never done: the verdict is schema-only and says so; (3) and (4) do not apply, this command never edits or posts.
|
|
28
28
|
|
|
29
29
|
## 1. Resolve the subject
|
|
30
30
|
|
|
@@ -20,7 +20,11 @@ Read `~/.claude/flow/CORE.opencode.md` first (shared rules: `FLOW.md` step 0, au
|
|
|
20
20
|
|
|
21
21
|
**Panel words are closed** — `mark`: `done` · `current` · `pending` · `wait` · `block` · `info`; `style`: `normal` · `dim` · `title` · `accent` · `ok` · `warn` · `error`. Anything else is dropped by the reader in silence: the panel still paints, and nobody is told.
|
|
22
22
|
|
|
23
|
-
Use when returning to a work after a break (next morning, another session)
|
|
23
|
+
Use when returning to a work after a break (next morning, another session) — including a work an
|
|
24
|
+
unattended run left waiting on a person (flow-core §2.1).
|
|
25
|
+
|
|
26
|
+
**In `unattended`** this command asks nothing: recap, name the next step, end. It is how a person
|
|
27
|
+
reads where the run got to, not how the run continues.
|
|
24
28
|
|
|
25
29
|
## 1. Detection
|
|
26
30
|
|
|
@@ -43,6 +47,7 @@ Last updated: <updated_at>
|
|
|
43
47
|
Notes: <meta.notes>
|
|
44
48
|
Cross-repo: <meta.related_repos entries not "done", as "repo: scope"; or "—">
|
|
45
49
|
Ticket thread: <new since <updated_at>: <n> comment(s) that change the work — detailed below · or "nothing new" · or "not read"> (line only in ticket mode, from §2.5)
|
|
50
|
+
Unattended: <from `.claude/work/stop.json` when its `work` is this one: `question` → the gate and the question, verbatim, with its options · `blocked` → the detail · `running` → "a run is working in <phase>, or died there, at <at>" · `done` → the MR/PR · `picked_up` → omit the line> (line only when that file names this work)
|
|
46
51
|
```
|
|
47
52
|
|
|
48
53
|
- `MR/PRs:` and `Waves:` from `meta.json.mrs` (same format as `/flow-work-status §2`). `∥` = can run in parallel, `→` = waits for the previous wave to merge.
|
|
@@ -84,6 +89,7 @@ With what is new (there is already work on disk):
|
|
|
84
89
|
|
|
85
90
|
## 4. Next step
|
|
86
91
|
|
|
92
|
+
- A pending `question` in the `Unattended:` line → say that the next phase will ask it first (flow-core §2.1); the next step itself comes from `phase` as always.
|
|
87
93
|
- Suggest the concrete command from `phase` and `size`. Interrupted phase (e.g. `build` with an empty artifact) → suggest rerunning it with `/flow-feat-build` or `/flow-bug-fix`.
|
|
88
94
|
- §2.5 found something new → let it inform the suggestion (a contract that arrived after you left may mean rerunning `design` rather than continuing `build`); say why in one line.
|
|
89
95
|
- `meta.json.related_repos` entries not `done` → remind that a **sibling repo still has a pending part** (`<repo>: <scope>`, plus `contract not handed over` when its `contract_handoff` is `pending`); suggest `/flow-feat-start <TICKET>` there. flow only reminds; it never scans or touches the other repo.
|
package/package.json
CHANGED
|
@@ -1,6 +1,6 @@
|
|
|
1
1
|
{
|
|
2
2
|
"name": "flow-workflows",
|
|
3
|
-
"version": "0.
|
|
3
|
+
"version": "0.82.0",
|
|
4
4
|
"description": "Guided feat/bug workflows (start→ship), post-deploy monitoring and multi-agent code review, stack-agnostic. Each repo uses a shared FLOW.md plus optional harness overlay.",
|
|
5
5
|
"type": "module",
|
|
6
6
|
"bin": {
|
|
@@ -3,7 +3,7 @@
|
|
|
3
3
|
"name": "flow",
|
|
4
4
|
"displayName": "Flow — guided dev workflows",
|
|
5
5
|
"description": "Guided feat/bug workflows (start→ship), post-deploy monitoring and multi-agent code review, stack-agnostic. Each repo uses a shared FLOW.md plus optional harness overlay.",
|
|
6
|
-
"version": "0.
|
|
6
|
+
"version": "0.82.0",
|
|
7
7
|
"author": {
|
|
8
8
|
"name": "Alberto Vioque",
|
|
9
9
|
"email": "mashware@gmail.com"
|
|
@@ -5,6 +5,28 @@ plugin and is what `/flow:news` reads to show you what changed since your previo
|
|
|
5
5
|
|
|
6
6
|
The canonical, richest notes live in the [GitHub Releases](https://github.com/mashware/flow-workflows/releases).
|
|
7
7
|
|
|
8
|
+
## v0.82.0 — Nobody at the terminal, and the run still gets somewhere · 2026-09-29
|
|
9
|
+
|
|
10
|
+
**In short**
|
|
11
|
+
- **A new autonomy mode, `unattended`, for runs nobody watches** — a script, a container, a headless
|
|
12
|
+
CLI. It behaves like `auto` and never asks. Before, such a run stopped for good at the first hard
|
|
13
|
+
gate: the brief, a serious review finding, `ship`, or any question left in prose for nobody.
|
|
14
|
+
- **It goes as far as it can without a person, and says exactly where it needs one.** The business
|
|
15
|
+
brief is written down and travels in the MR/PR body. `ship` opens a **draft** MR/PR only with
|
|
16
|
+
`autonomy.unattended_ship: draft` — never merged, never marked ready, never force-pushed. A
|
|
17
|
+
migration, serious review findings, a product decision or a doubt about the ticket end the run with
|
|
18
|
+
the question and its options in `.claude/work/stop.json`; a checkout or ticket the run cannot use
|
|
19
|
+
ends it as `blocked`. Nothing that protects the repo is approved in silence.
|
|
20
|
+
- **A runner reads one file.** `.claude/work/stop.json` says `running` while a phase works (left
|
|
21
|
+
behind = the run crashed), then `question`, `blocked` or `done` with the MR/PR. While a question is
|
|
22
|
+
pending no new unattended run goes past it. A person picks the work up in their own session:
|
|
23
|
+
`/flow:work:resume` shows the question, the next phase asks it first, and the file then says
|
|
24
|
+
`picked_up`. Keep `unattended` in a FLOW file only the runner's environment reads.
|
|
25
|
+
- **One MR/PR per run**, and only when written by hand: `unattended` is never inferred from a
|
|
26
|
+
headless harness, `/flow:init` does not offer it, and `manual`, `guided` and `auto` are unchanged.
|
|
27
|
+
|
|
28
|
+
Asked for in #168. Continuing a run automatically from an answer is not in this release.
|
|
29
|
+
|
|
8
30
|
## v0.81.0 — The review stopped contradicting itself · 2026-09-26
|
|
9
31
|
|
|
10
32
|
**In short**
|
|
@@ -89,7 +89,7 @@ Branch and Pull/Merge Request conventions.
|
|
|
89
89
|
## autonomy
|
|
90
90
|
How much the flow advances on its own vs. stopping to ask you.
|
|
91
91
|
|
|
92
|
-
- `mode:` # `manual` (default) | `guided` | `auto`. Empty = `manual`.
|
|
92
|
+
- `mode:` # `manual` (default) | `guided` | `auto` | `unattended`. Empty = `manual`.
|
|
93
93
|
# manual — every phase stops at each decision point and, at the end, proposes the
|
|
94
94
|
# next command as a one-click confirmation (you accept to advance, it is
|
|
95
95
|
# never typed for you and never runs without your confirmation).
|
|
@@ -99,7 +99,12 @@ How much the flow advances on its own vs. stopping to ask you.
|
|
|
99
99
|
# automatically (no advance confirmation).
|
|
100
100
|
# auto — as guided, plus auto-resolves the remaining decision points with sensible
|
|
101
101
|
# (recorded) defaults, chaining phases without pausing.
|
|
102
|
-
#
|
|
102
|
+
# unattended — as auto, for a run nobody watches (script, container, headless CLI):
|
|
103
|
+
# never asks; each hard gate is recorded, becomes a draft MR/PR, or ends
|
|
104
|
+
# the run with a question in .claude/work/stop.json for a runner to
|
|
105
|
+
# post and a person to pick up. One MR/PR per run. Only when set explicitly —
|
|
106
|
+
# in a FLOW file only the runner's environment reads.
|
|
107
|
+
# HARD GATES stop and ask in manual/guided/auto, no exceptions: any push or MR/PR (ship),
|
|
103
108
|
# branch creation with an ambiguous base, DB schema changes/migrations, a review
|
|
104
109
|
# with high-severity findings, and the business brief confirmed just before the
|
|
105
110
|
# first edit in `build`/`fix`.
|
|
@@ -107,6 +112,10 @@ How much the flow advances on its own vs. stopping to ask you.
|
|
|
107
112
|
# (panels, challengers, how many reviewers), WIP commits, continuing a train when
|
|
108
113
|
# the train's next MR/PR in `guided`/`auto`, size confirmation, and anything already decided
|
|
109
114
|
# and recorded — only new contradicting evidence reopens a settled decision.
|
|
115
|
+
- `unattended_ship:` # What `ship` does in `unattended`: `draft` | `stop`. Empty = `stop`.
|
|
116
|
+
# draft — push the work branch and open the MR/PR as draft; never merge, never ready.
|
|
117
|
+
# stop — nothing is pushed; ship ends the run with a question in stop.json.
|
|
118
|
+
# Read only when mode is `unattended`.
|
|
110
119
|
|
|
111
120
|
## quality
|
|
112
121
|
Repo commands for quality gates, whatever the stack — `make test`, `./gradlew test`, `dotnet test`,
|
|
@@ -256,7 +265,7 @@ subagent ignores the value and the step says so once, in one line.
|
|
|
256
265
|
# brings back status plus evidence — job names, exit codes, log tail, run URL — and never
|
|
257
266
|
# rules on it: triage and the merge verdict stay on the main thread whatever this says.
|
|
258
267
|
# Empty = falls back to `agents`, then to what the command runs on, and the dispatch says so
|
|
259
|
-
# in one line — an
|
|
268
|
+
# in one line — an unwatched wait is the one place an unpriced default runs longest.
|
|
260
269
|
|
|
261
270
|
**Three keys, one per kind of subagent — not one per kind of step.** There used to be five —
|
|
262
271
|
`study`, `code`, `test`, `review` — named after phases. They promised a granularity no harness
|