handback-digest
Draft one choices-ledger entry per choice in a lane handback, in the ledger's own ELI5 entry format, as the starting point for the orchestrator's…
Skill by threecee
Install
Works with Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf. Read the repository before you give any tool access to your data.
npx agentmarket install skill/threecee-handback-digest
Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.
Get the weekly diff
The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.
About
Use cases: security
Reviews
Write a reviewNo reviews yet. Installed it? Reviews come from verified installers only. How reviews work
Permissions
What this skill can do and what it asks for, from its SKILL.md. Checked Oct 25, 2026.
- Tools
- No tools found
- Secrets
- None asked for
- Install scripts
- None found
Network hosts it talks to 1
codeload.github.com· found in install command
Files it touches
~/.claude/skills/handback-digest
What's inside
- Skill name
handback-digest- Description
- Draft one choices-ledger entry per choice in a lane handback, in the ledger's own ELI5 entry format, as the starting point for the orchestrator's audit-choices pass. Use after a lane files its handback and before the orchestrator writes verdicts into the train's choices ledger. The draft is never the verdict.
- License
- MIT
- Instructions
- 65 lines in SKILL.md · view on GitHub
Files (1)
- SKILL.md
SKILL.md (first lines)
# Handback Digest
The handback-digest command reads a lane's handback (the report the launcher
places at `LANE_RESULT_PATH` — `<lane>.<run-id>.result.md`,
`harness/run-lifecycle.md` §3 — written to the fields in
`harness/report-schema.md`, plus the `<lane>-choices.md` sidecar its
`choices_sidecar:` field names whenever `choices:` is non-empty) and drafts
one structured entry per choice: headline, a
provisional verdict, confidence, and one concrete end-to-end scenario, with
terms defined in place. The rules live in `harness/bulk-read-contract.md`;
this skill is the invocation and the hand-over to the audit.
## Invocation
```sh
{{HANDBACK_DIGEST_CMD}} <path-to-handback> [--diff-stat]
```
- `--diff-stat` appends `git diff --stat` for the lane's tree as extra
context for the draft; it adds file names and counts, not content.
- Same input boundary as the other shunts (repository + lane scratchpad;
exit 2 refused, exit 3 every worker failed, nothing echoed);
`{{ALLOW_EXTERNAL_FLAG}}` under the same condition.
- Invoked deliberately, never via a hook block: a handback is short by
contract, so the digest is a drafting aid, not a context saver.
## What the draft must preserve
Score the draft against the handback before using it (contract §6): every
exact identifier — lane name, head SHA, issue numbers, gate names and their
red/green result, test names, the `proof` command — and every negation
("not run", "skipped", "deselected") must survive verbatim. A digest that
turns "gate X not run" into "gates green" is misstated, not merely
incomplete; discard it and narrow the question.
## Hand-over to the audit
The output is a DRAFT for the orchestrator's `skills/audit-choices` pass,
never the final verdict (contract §3 rule 7):
1. `audit-choices` traces the handback **and** the diff itself; the digest's
list is a starting point, not the boundary — workers under-report as
agents do.
2. Each draft entry carries the same fields the ledger stores (headline,
walked scenario, gap, reach, verdict, confidence — see
`interpretation/choices-ledger-README.md`) so the auditor edits an entry
rather than re-typing it; the auditor writes the verdict and the entry
into the train's ledger under the stable choice ID
(`interpretation/choices-ledger-README.md` §2).
3. Never treat the digest as already reviewed, never paste it into the ledger
unedited, and never let it edit any file (it does not).
## Escape hatch
`{{SHUNT_DISABLE}}` disables the bounded-read hooks for the session; it does
not gate this command, which is invoked on purpose. When every worker is
down, audit the handback directly — it is short by contract.
## Parameters
`{{HANDBACK_DIGEST_CMD}}`, `{{ALLOW_EXTERNAL_FLAG}}` and `{{SHUNT_DISABLE}}`
are bound in the repository's operations doc from
`harness/bulk-read-contract.md` §4.From threecee/software-factory-kit, MIT.
More from threecee/software-factory-kit
- 0.563score
code-review
threecee
Review a current diff for correctness, stale references, ownership, test quality, unnecessary compatibility, and cleanup.
SkillBFreshness: B★ 2Claude Code +9 - 0.563score
compare-screenshots
threecee
Compare screenshots to judge which image is less wrong, not to match a baseline.
SkillBFreshness: B★ 2Claude Code +9 - 0.563score
refactor-clean
threecee
Refactor toward one clear owner per concept instead of adding sediment.
SkillBFreshness: B★ 2Claude Code +9 - 0.559score
implement-spec
threecee
Implement an approved spec through reviewable passes.
SkillBFreshness: B★ 2Claude Code +9 - 0.559score
audit-choices
threecee
Audit decisions an implementing agent made beyond the brief, without changing code.
SkillBFreshness: B★ 2Claude Code +9 - 0.559score
independent-lane-review
threecee
Run a fresh-context, multi-model adversarial review of a lane before train assembly or on demand.
SkillBFreshness: B★ 2Claude Code +9 - 0.559score
finishing-a-development-branch
threecee
Use after implementation and tests are complete to choose and safely execute merge, pull-request, keep, or discard handling while respecting normal, linked…
SkillBFreshness: B★ 2Claude Code +9 - 0.559score
brainstorming
threecee
Explore intent and produce an approved design before unbriefed creative implementation.
SkillBFreshness: B★ 2Claude Code +9 - 0.557score
bulk-reader
threecee
Answer a narrow question over one or more large repository files without reading their raw content into context.
SkillBFreshness: B★ 2Claude Code +9 - 0.557score
log-triage
threecee
Extract failing tests, assertion messages, tracebacks, and red gate results from a large run or verify log without reading its full text.
SkillBFreshness: B★ 2Claude Code +9 - 0.557score
claude
threecee
Use Claude Code as an independent `claude -p` subagent when the user explicitly asks for Claude, wants a second-agent opinion from Claude, or asks to delegate…
SkillBFreshness: B★ 2Claude Code +9
Signals
- Stars
- 2
- Forks
- 0
- Open issues
- 0
- Last commit (this folder)
- Sep 6, 2026 (1 month ago)
From GitHub (fetched Oct 24, 2026).
Details
- License
- MIT
- Repository
- github.com/threecee/software-factory-kit
- First seen
- Oct 24, 2026
- Last checked
- Oct 24, 2026
- Source
- GitHubgithub-graphql
More from threecee
- 0.563score
code-review
threecee
Review a current diff for correctness, stale references, ownership, test quality, unnecessary compatibility, and cleanup.
SkillBFreshness: B★ 2Claude Code +9 - 0.563score
compare-screenshots
threecee
Compare screenshots to judge which image is less wrong, not to match a baseline.
SkillBFreshness: B★ 2Claude Code +9 - 0.563score
refactor-clean
threecee
Refactor toward one clear owner per concept instead of adding sediment.
SkillBFreshness: B★ 2Claude Code +9 - 0.559score
implement-spec
threecee
Implement an approved spec through reviewable passes.
SkillBFreshness: B★ 2Claude Code +9 - 0.559score
audit-choices
threecee
Audit decisions an implementing agent made beyond the brief, without changing code.
SkillBFreshness: B★ 2Claude Code +9 - 0.559score
independent-lane-review
threecee
Run a fresh-context, multi-model adversarial review of a lane before train assembly or on demand.
SkillBFreshness: B★ 2Claude Code +9
Cite this page
handback-digest skill. Agent Mart. https://agentmart.dev/skills/threecee-handback-digest (last updated 2026-10-24)