BACKS Build Flow
One integration owner carries the request from context to verified delivery.
Specialist skills return evidence to that owner; they do not restart the workflow.
For a simple answer with no build or investigation, use only the relevant steps.
Model agnostic: resolve roles by required capability and actual availability.
The same capable model may fill multiple roles in separate executions. Required
independent review uses a fresh context that did not participate in authoring or
inherit the author's conversation. No specific model, provider, family, or model
diversity is required. Record the selected route as provenance, not authority.
Establish enough context to act
- Preserve the user's objective, constraints, existing authorization, and requested
outcome. Choose a mode: build/fix, read-only review, investigate, or
resume/deliver. A review does not authorize edits, an install, or deployment.
- Reuse a current task packet. Otherwise use
context-engineer to verify the target, governing
instructions, affected caller/consumer path, existing work, and acceptance check.
building-with-backs owns checkout and promotion
resolution. Pin the pre-change
base_sha; update candidate_sha as work commits.
- Apply operator-essence for the current operator:
load the relevant awareness slice, or run its missing-profile insight bootstrap
once. Qualitative observations and quantitative evidence guide communication
and verification; they do not grant authority or hardcode a persona. Mark sparse
evidence provisional and continue useful authorized work. Use
intent-compiler only for ambiguous intent;
operator-intent-deduction owns deciding
whether a material unresolved choice really needs the operator.
- Start execution when you can name the target, next concrete action, constraints,
and an observable acceptance check. Further reading must resolve a named
question that could change that action. Do not repeatedly map the whole repo.
Route skills once, with an output and an exit
Load actual relevant skill bodies; a name or receipt alone is not application.
Reuse bodies and evidence still present and current in context. After compaction,
reload only what the next action needs. For each selected specialist, state its
purpose, expected evidence, and exit condition; read deeper references on demand.
One current task packet tracks ownership, state, and the next action
(packet and handoff format). Avoid duplicate ledgers.
For a Codex run that requests native same-thread continuation, bind that packet at
each durable checkpoint with python hooks/codex_stop_continuation.py bind-task-packet RUN_ID < task-packet.json. The packet is the producer; the Stop
adapter is the callback owner. Rebind after recording real progress and cumulative
usage, never from an inferred or reset counter.
For adoption, absorb owns capability mapping and native
integration. rea-investigation supplies discovery
or verification evidence when relevant; it returns to its caller without
recursively starting Absorb. Ordinary source work uses source and tests directly.
Execute the bounded RED → GREEN loop
Bind bounded-loops before retries: one mission_id,
deadline, attempt/round allowance, and remaining resource budget. Child workers,
model changes, repair findings, and review rounds spend that same episode allowance.
Carry the same mission and evidence through implementation, review, and landing.
A self-imposed deadline or attempt count ends the bounded episode, not the authorized
mission: checkpoint at a safe boundary, preserve the counters, and continue with the
next useful action only after binding bounded-loops again
with a freshly declared episode allowance. The task packet retains the exhausted
episode's counters alongside the new allowance; continuation never resets or overwrites
prior usage. Respect actual provider limits, explicit operator constraints,
cancellation, and required review.
- RED: red-first establishes a real failing behavior
check before the fix. Preserve its sealed file and baseline identity. For a
review or research task, use its evidence contract without manufacturing edits.
- Build: implement the smallest complete change at the owning seam; update
affected consumers and compatibility surfaces. Preserve unrelated work.
- Lint and test: sniper-testing derives scope from
base_sha through the current candidate, including relevant pending work.
Execute the repository's applicable lint, format-check, typecheck, and behavior
tests. Record exact commands, revision, results, and unavailable or inapplicable
checks. A clean working tree does not erase committed scope.
- Quality: clean-code-gauntlet uses that same
scope for measured quality, isolated mutation, and independent
review. Add supplemental tests for new gaps without rewriting the sealed RED.
- Respond to evidence: fix a proven finding and rerun only affected checks.
Repeated identical failures require a changed hypothesis or tactic. Report actual
blockers and preserve evidence. When an episode allowance is exhausted, checkpoint
it before continuing the mission; do not lower the checks, reset old counters, or
claim a worker is active when it is not.
Use the repository's required aggregate landing gates once the candidate settles.
Report actual capabilities of this host; do not invent tools, tests, or reviews.
Finish the requested outcome
Continue through authorized packaging, commit, publication, and verification when
those are part of the request. For external approval that is genuinely required,
first prepare the complete reviewable result. Keep the user informed of findings,
current action, evidence, and where to inspect live work; follow their cadence.
Distinguish requested task delivered from the invariant floor's PROVEN
capability status. A completed local patch or read-only review does not imply live
deployment. Failed or unavailable required gates remain visible: checkpoint
STILL-BUILDING with the exact blocker, actual worker state, and next action.
In read-only mode return the packet in the response or an already-authorized
external task record; do not create a repository ledger or essence file implicitly.
1---2name: backs-build-flow3description: "Use when building, fixing, reviewing, resuming, or delivering work that requires coordinating context, skills, and verification without losing execution momentum."4license: MIT5metadata:6 source: https://skillmd.com/skills/tcuzzo/backs-build-flow7---89# BACKS Build Flow1011One integration owner carries the request from context to verified delivery.12Specialist skills return evidence to that owner; they do not restart the workflow.13For a simple answer with no build or investigation, use only the relevant steps.1415**Model agnostic:** resolve roles by required capability and actual availability.16The same capable model may fill multiple roles in separate executions. Required17independent review uses a fresh context that did not participate in authoring or18inherit the author's conversation. No specific model, provider, family, or model19diversity is required. Record the selected route as provenance, not authority.2021## Establish enough context to act22231. Preserve the user's objective, constraints, existing authorization, and requested24 outcome. Choose a mode: **build/fix**, **read-only review**, **investigate**, or25 **resume/deliver**. A review does not authorize edits, an install, or deployment.262. Reuse a current task packet. Otherwise use27 [context-engineer](../context-engineer/SKILL.md) to verify the target, governing28 instructions, affected caller/consumer path, existing work, and acceptance check.29 [building-with-backs](../building-with-backs/SKILL.md) owns checkout and promotion30 resolution. Pin the pre-change `base_sha`; update `candidate_sha` as work commits.313. Apply [operator-essence](../operator-essence/SKILL.md) for the current operator:32 load the relevant awareness slice, or run its missing-profile insight bootstrap33 once. Qualitative observations and quantitative evidence guide communication34 and verification; they do not grant authority or hardcode a persona. Mark sparse35 evidence provisional and continue useful authorized work. Use36 [intent-compiler](../intent-compiler/SKILL.md) only for ambiguous intent;37 [operator-intent-deduction](../operator-intent-deduction/SKILL.md) owns deciding38 whether a material unresolved choice really needs the operator.394. **Start execution when you can name the target, next concrete action, constraints,40 and an observable acceptance check.** Further reading must resolve a named41 question that could change that action. Do not repeatedly map the whole repo.4243## Route skills once, with an output and an exit4445Load actual relevant skill bodies; a name or receipt alone is not application.46Reuse bodies and evidence still present and current in context. After compaction,47reload only what the next action needs. For each selected specialist, state its48purpose, expected evidence, and exit condition; read deeper references on demand.49One current task packet tracks ownership, state, and the next action50([packet and handoff format](references/task-packet.md)). Avoid duplicate ledgers.51For a Codex run that requests native same-thread continuation, bind that packet at52each durable checkpoint with `python hooks/codex_stop_continuation.py bind-task-packet RUN_ID < task-packet.json`. The packet is the producer; the Stop53adapter is the callback owner. Rebind after recording real progress and cumulative54usage, never from an inferred or reset counter.5556For adoption, [absorb](../absorb/SKILL.md) owns capability mapping and native57integration. [rea-investigation](../rea-investigation/SKILL.md) supplies discovery58or verification evidence when relevant; it returns to its caller without59recursively starting Absorb. Ordinary source work uses source and tests directly.6061## Execute the bounded RED → GREEN loop6263Bind [bounded-loops](../bounded-loops/SKILL.md) before retries: one `mission_id`,64deadline, attempt/round allowance, and remaining resource budget. Child workers,65model changes, repair findings, and review rounds spend that same episode allowance.66Carry the same mission and evidence through implementation, review, and landing.67A self-imposed deadline or attempt count ends the bounded episode, not the authorized68mission: checkpoint at a safe boundary, preserve the counters, and continue with the69next useful action only after binding [bounded-loops](../bounded-loops/SKILL.md) again70with a freshly declared episode allowance. The task packet retains the exhausted71episode's counters alongside the new allowance; continuation never resets or overwrites72prior usage. Respect actual provider limits, explicit operator constraints,73cancellation, and required review.74751. **RED:** [red-first](../red-first/SKILL.md) establishes a real failing behavior76 check before the fix. Preserve its sealed file and baseline identity. For a77 review or research task, use its evidence contract without manufacturing edits.782. **Build:** implement the smallest complete change at the owning seam; update79 affected consumers and compatibility surfaces. Preserve unrelated work.803. **Lint and test:** [sniper-testing](../sniper-testing/SKILL.md) derives scope from81 `base_sha` through the current candidate, including relevant pending work.82 Execute the repository's applicable lint, format-check, typecheck, and behavior83 tests. Record exact commands, revision, results, and unavailable or inapplicable84 checks. A clean working tree does not erase committed scope.854. **Quality:** [clean-code-gauntlet](../clean-code-gauntlet/SKILL.md) uses that same86 scope for measured quality, isolated mutation, and independent87 review. Add supplemental tests for new gaps without rewriting the sealed RED.885. **Respond to evidence:** fix a proven finding and rerun only affected checks.89 Repeated identical failures require a changed hypothesis or tactic. Report actual90 blockers and preserve evidence. When an episode allowance is exhausted, checkpoint91 it before continuing the mission; do not lower the checks, reset old counters, or92 claim a worker is active when it is not.9394Use the repository's required aggregate landing gates once the candidate settles.95Report actual capabilities of this host; do not invent tools, tests, or reviews.9697## Finish the requested outcome9899Continue through authorized packaging, commit, publication, and verification when100those are part of the request. For external approval that is genuinely required,101first prepare the complete reviewable result. Keep the user informed of findings,102current action, evidence, and where to inspect live work; follow their cadence.103104Distinguish **requested task delivered** from the invariant floor's **PROVEN**105capability status. A completed local patch or read-only review does not imply live106deployment. Failed or unavailable required gates remain visible: checkpoint107**STILL-BUILDING** with the exact blocker, actual worker state, and next action.108In read-only mode return the packet in the response or an already-authorized109external task record; do not create a repository ledger or essence file implicitly.