Independent testing will determine whether PhysiClaw can reliably operate physical iPhones for practical computer-use agent workflows.
state: expiredheat: lowuncertainty: highknownscott: lowcomputer-use agent-harnesses mobile-agentsPhysiClaw
What is this?
PhysiClaw is presented in a Show HN post as an AI agent that physically operates an iPhone, with its README describing camera-based “eyes” and an incompletely quoted physical-control mechanism. The supplied material does not identify its creators, provide implementation details, or report independent test results. General mobile-testing snippets support the premise that real-device testing exposes hardware-, gesture-, and environment-specific failures missed by simulators, so PhysiClaw’s practical reliability remains an open empirical question.
Why it matters to Scott
The radar already tracks the same real-iPhone-control validation question in “Independent use will determine whether mobile-harness reliably controls real iOS” and the Sidetap iPhone MCP case. PhysiClaw is another unvalidated hands-and-eyes harness consistent with Scott’s model-plus-harness and capability-audit positions, but the supplied evidence offers no implementation detail or independent results that extend or challenge them.
ip:concept.model-plus-harness-benchmark-unitip:concept.agent-hands-and-eyesip:concept.capability-auditradar:mobile-harness-cross-platform-controlradar:sidetap-iphone-mcp-windowsradar:concept.mobile-agentsradar:concept.computer-use-agents
queries asked of Scott's wikis
- physical-device computer-use agent reliability
- vision-and-actuation agent harnesses
- mobile agent evaluation and pass^k
- closed-loop verification for GUI agents
- real-device versus simulator testing
- safe permissions and recovery for autonomous phone control
Measured heat
no measured readings yet — the hourly heat pass fills this in
How the heat travelled
no chain yet — the hourly chain pass fills this in
Evidence (3) — ⭐ canonical anchor
Interpretation history
2026-08-20T18:31:49Z
Minor engagement adds no independent testing, implementation detail, or reliability evidence for PhysiClaw. As a low-relevance duplicate of an already tracked mobile-agent pattern with no further development, this episode has faded beyond its useful horizon.
2026-08-18T17:38:44Z
A second independently released phone-control harness makes physical mobile-agent actuation a recurring implementation pattern rather than a one-off concept. It does not test PhysiClaw or establish practical reliability, so the core validation question remains open.
2026-08-18T15:24:03Z
evidence attached: hn.story.49346652 — Another released phone-control harness provides directly relevant corroborating evidence for practical agent operation of mobile devices.
2026-08-18T13:45:57Z
The only change is negligible engagement without comments, independent testing, implementation detail, or demonstrated reliability. PhysiClaw remains an unvalidated example of an already tracked physical-phone agent pattern.
2026-08-18T13:38:18Z
grounded: known/low — The radar already tracks the same real-iPhone-control validation question in “Independent use will determine whether mobile-harness reliably controls real iOS”
2026-08-18T13:35:42Z
origin walked (codex/luna, conf 0.94): anchor hn.story.49345058 -> echo.github.84e3e10748 by Qiao Qian (qiaoqian / physiclaw)
2026-08-18T13:34:18Z
case created — The repository is a concrete mobile computer-use artifact with practical reliability still unvalidated.
Decision trace
- 08-21 04:31expireMinor engagement adds no independent testing, implementation detail, or reliability evidence for PhysiClaw. As a low-relevance duplicate of an already tracked mobile-agent pattern with no further deve
- 08-21 04:31alert_silentThe new delta is only modest engagement without substantive evidence; there is no event or result that should interrupt Scott or justify a short hold.
- 08-21 04:31alert_routeThe new delta is only modest engagement without substantive evidence; there is no event or result that should interrupt Scott or justify a short hold.
- 08-19 03:38repriceA second independently released phone-control harness makes physical mobile-agent actuation a recurring implementation pattern rather than a one-off concept. It does not test PhysiClaw or establish pr
- 08-19 03:38alert_silentThe additional harness is ecosystem corroboration, not independent testing or a demonstrated capability result for PhysiClaw; it can wait for a normal briefing until benchmarks, implementation details
- 08-19 03:38alert_routeThe additional harness is ecosystem corroboration, not independent testing or a demonstrated capability result for PhysiClaw; it can wait for a normal briefing until benchmarks, implementation details
- 08-19 01:25alert_silentPhone-harness is another unvalidated phone-control project with no supplied implementation details, independent testing, or demonstrated reliability that materially extends the existing PhysiClaw and
- 08-19 01:25alert_routePhone-harness is another unvalidated phone-control project with no supplied implementation details, independent testing, or demonstrated reliability that materially extends the existing PhysiClaw and
- 08-19 01:24attachAnother released phone-control harness provides directly relevant corroborating evidence for practical agent operation of mobile devices.
- 08-19 01:23propose_attachAnother released phone-control harness provides directly relevant corroborating evidence for practical agent operation of mobile devices.
- 08-18 23:45repriceThe only change is negligible engagement without comments, independent testing, implementation detail, or demonstrated reliability. PhysiClaw remains an unvalidated example of an already tracked physi
- 08-18 23:45alert_silentNo consequential new fact has emerged beyond the original project claim, so this can wait for independent results or a substantive implementation release.
- 08-18 23:45alert_routeNo consequential new fact has emerged beyond the original project claim, so this can wait for independent results or a substantive implementation release.
- 08-18 23:43alert_silentPhysiClaw is a newly public physical iPhone-control harness, but the supplied evidence contains only its high-level camera-and-robotic-arm concept, with no implementation detail, release artifact, ind
- 08-18 23:43alert_routePhysiClaw is a newly public physical iPhone-control harness, but the supplied evidence contains only its high-level camera-and-robotic-arm concept, with no implementation detail, release artifact, ind
- 08-18 23:38groundThe radar already tracks the same real-iPhone-control validation question in “Independent use will determine whether mobile-harness reliably controls real iOS” and the Sidetap iPhone MCP case. PhysiCl
- 08-18 23:35promote_anchororigin walk conf 0.94
- 08-18 23:34createThe repository is a concrete mobile computer-use artifact with practical reliability still unvalidated.