Add human factors problem document#9
Merged
Merged
Conversation
Cover the soft concerns around autonomous agents: domain ownership, role shift from author to supervisor, review/intent fatigue, contributor motivation, and job security. These are explicitly absent from the existing problem documents which focus on technical and process concerns. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Contributor
|
Is this related to #7 ? |
arewm
approved these changes
Mar 11, 2026
|
|
||
| This is still fatiguing, in different ways: | ||
|
|
||
| - **Volume.** Agents can generate and process changes faster than humans can evaluate whether the results match what they wanted. The bottleneck moves from "review this diff" to "verify this outcome aligns with what I meant." |
Contributor
There was a problem hiding this comment.
Human code review is unsustainable for these agentic workflows. But human intent review is also unsustainable as you imply. I think that the ultimate intention for fullsend is for this to not be required as well. How might would this problem document change under those conditions?
| - **Agents as force multipliers, not replacements.** Design workflows where agents handle toil so humans can focus on harder, more interesting problems. But verify that this actually happens in practice — it's easy for "focus on harder problems" to quietly become "there are no problems left for you." | ||
| - **Rotation and growth.** If domain experts risk skill atrophy, create deliberate opportunities for hands-on work — spikes, experiments, prototypes that agents don't touch. | ||
| - **Transparent metrics.** Track not just agent effectiveness but human engagement. If humans are rubber-stamping intent approvals on guarded paths, the system is failing even if the code is correct. | ||
| - **Contributor pathways.** Explicitly design how new contributors enter the project when agents handle the easy on-ramps. Mentorship, pairing, or reserved "human-first" areas could help. |
Contributor
There was a problem hiding this comment.
Some communities mark specific issues as non-AI issues to provide easy opportunities for new contributors.
ralphbean
reviewed
Mar 11, 2026
|
|
||
| This is still fatiguing, in different ways: | ||
|
|
||
| - **Volume.** Agents can generate and process changes faster than humans can evaluate whether the results match what they wanted. The bottleneck moves from "review this diff" to "verify this outcome aligns with what I meant." |
Contributor
There was a problem hiding this comment.
Jeremy Eder just published a post that touches on this topic. See the section titled The meeting in https://ambient-code.ai/2026/03/10/structure-dictates-behavior-golden-signals-for-agentic-development-teams/
chmeliik
reviewed
Mar 12, 2026
Comment on lines
+21
to
+25
| The vision document says "humans set direction, agents execute." For many contributors, this describes a less satisfying way to participate. Writing code, debugging, and shipping are core to why people contribute to open-source projects. Supervising agents is a fundamentally different activity. | ||
|
|
||
| **What changes:** | ||
|
|
||
| - **Creative work decreases.** Design and architecture remain human, but the hands-on problem-solving that many contributors enjoy moves to agents. |
waynesun09
added a commit
that referenced
this pull request
Apr 12, 2026
Seven specialized agents for working on the fullsend project: - fullsend-architect (opus): architectural coherence guardian; knows all ADRs, five execution layers, story dependencies, repo-as-coordinator invariant - go-developer (sonnet): CLI specialist; forge abstraction, layered config, multi-role GitHub App model, known gaps in PR #132 - doc-architect (sonnet): problem doc and ADR writer; design-exploration conventions, org-agnostic authoring rules - stage-prompt-designer (opus): designs/reviews stage agent prompts; triage/implement/review/fix constraints, injection surface rules, known failure modes from live operation (Issues #4, #5, #010a) - security-reviewer (opus): applies fullsend threat model; prompt injection, ADR 0017 credential isolation, sandbox integrity, workflow file protection - workflow-engineer (sonnet): GitHub Actions and dispatch layer; label state machine, slash commands, concurrency groups, fixes for Issues #1 #003b #4 #5 #7 #9 #010a - e2e-integrator (opus): full flow tracing; integration gap analysis, demo readiness checklist, sprint prioritization across stories Also adds .claude/AGENTS.md with usage guide and team composition patterns.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
4 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Cover the soft concerns around autonomous agents: domain ownership, role shift from author to supervisor, review/intent fatigue, contributor motivation, and job security. These are explicitly absent from the existing problem documents which focus on technical and process concerns.