Standard 1.5: Outcome-oriented execution and bounded supporting work

[artyomboyko]

Summary

Publish Agent Handoff Standard 1.5 with outcome-oriented execution, bounded supporting work, explicit authorization boundaries, outcome-based stage and handoff rules, and progress-stalled recovery.

Why

Standard 1.4 constrained disproportionate security and evidence work, but the general workflow still allowed a failed smoke wrapper, test harness, CI layer, evidence collector, release helper, or other supporting tool to become its own stable stage, handoff, approval gate, and completion target.

That gap could create repeated process loops without advancing the Issue's primary outcome. Standard 1.5 closes it by requiring a primary outcome, smallest acceptance proof, and execution envelope; treating post-fix verification as part of the current work item; and limiting stages, handoffs, and new owner decisions to explicit outcome or approval boundaries.

Owner review added one final safety clarification: the execution envelope records existing authorization and cannot be used by an agent to create or widen its own authority.

Related issue

Closes #16

Agent handoff

• Agent ID: steady-path-8f2c
• Run ID: 20260726-issue-16-8f2c
• Work claim exists: yes
• Stage or final result comment exists: yes
• Current handoff: ai/handoffs/2026-07-26_issue-16_pr-17_run-20260726-issue-16-8f2c.md

Change type

• Bug fix
• Standard change
• Documentation
• Template change
• Refactoring
• Research
• Other

Checks

• Related Issue or PR is linked.
• Required files exist.
• Links work.
• Forms are valid when changed.
• Required stage or final result comment exists.
• Smoke tests were run or reason is documented.
• Primary outcome, smallest acceptance proof, and execution envelope are recorded; supporting work did not become an independent stage, handoff, completion target, or approval gate without crossing an explicit boundary.
• Security and evidence work did not block acceptance or expand scope without a verified High/Critical current-scope risk or an exactly cited mandatory requirement; other items remain non-blocking.
• Mandatory containerization decision was confirmed when relevant, or not applicable.
• Changed Compose configuration was rendered and checked, or reason and risk are documented.
• Automated GUI tests avoid position-dependent selectors, or manual/supervised checks are documented.

State updates

• ai/PROJECT_STATE.md updated, or not needed.
• ai/DECISIONS.md updated, or not needed.
• Handoff file added, or not needed.
• ai/handoffs/INDEX.md updated, or not needed.

Validation

• python scripts/check_agent_handoff.py — passed
• python -m py_compile scripts/check_agent_handoff.py — passed with temporary cache outside the repository
• git diff --check — passed
• Relative Markdown link validation — passed
• Active version scan — no stale metadata or current-workflow instructions
• Final tree SHA 6a6be490676f36131e1dc9ca92dbec44a4243983 matches the locally validated tree
• GitHub Actions 30198041516 — success

Compatibility

• The execution envelope records but does not grant or widen authority.
• Existing proportional-security rules and security baselines remain unchanged.
• Existing containerization decision gates remain unchanged.
• Existing GUI testing rules remain unchanged.
• Tool, platform, repository, and organization permission systems remain authoritative.
• No automated semantic progress scoring was added.

Risks

No verified blocking risk. Owner review accepted the outcome-oriented execution and approval-boundary semantics.

Publication

• Final reviewed PR HEAD: 897abc70625fa8ad464117389f3c3e28cc71080e
• Merge method: squash
• Tag v1.5 and the GitHub Release must be created only from the verified merge commit.