meta: release pressure-test run log (cold-read usability sweeps)

coilyco-ops commented

2026-07-01 20:24:03 +00:00

Owner

What this is

Coordination + trust artifact for the repeated cold-read release pressure tests: a dozen-ish fresh agent sessions, each simulating a cold user landing on this repo from the outside, each filing adoption blockers as issues labeled pressure-test. The runbook prompt lives on Kai's host (~/projects/ward-pressure-test-prompt.md).

Protocol for each run

Append exactly ONE comment here per run, in this shape and nothing more:

run: <n>                      (n = number of existing run comments + 1)
persona: <persona key played>
angle: <surface gone deep on>
filed: #a, #b, #c   (+1'd: #x, #y)
verdict: would this persona adopt today? yes/no - <top blocker, one line>

(Exception: drive-by-critic runs replace the verdict line with: verdict: does the front page survive a hostile thread? yes/no - <top-voted dunk, one line>.)

No finding details in this log. Later runs read ONLY this issue before their cold read (to pick an uncovered persona/angle and self-assign a run number); finding details here would contaminate their first impressions.

Exit condition (the release trust gate)

The sweep is done when two consecutive runs produce zero new adoption-fatal or major-friction findings and their verdicts are yes. At that point the pressure-test label is the launch burndown: fix adoption-fatal + major-friction, batch or defer polish, ship.

Known anchors (pre-existing - comment there instead of re-filing)

ward#414 - add github equivalents for forgejo-paired functionality
ward#208 - ward website
ward#190 - pressure test ward without ward or cli-guard on disk
ward#201 - "blind" openclaw / opencode agent pressure testing
ward#255 - launch writeup (doctrine: every demo shows a denial)
ward#195 - doctor pattern for onboarding onto the ward agent surface
ward#427 - lift the README length cap

Filed by Claude Code during the release pressure-test setup session.

## What this is Coordination + trust artifact for the repeated cold-read release pressure tests: a dozen-ish fresh agent sessions, each simulating a cold user landing on this repo from the outside, each filing adoption blockers as issues labeled `pressure-test`. The runbook prompt lives on Kai's host (`~/projects/ward-pressure-test-prompt.md`). ## Protocol for each run Append exactly ONE comment here per run, in this shape and nothing more: ``` run: <n> (n = number of existing run comments + 1) persona: <persona key played> angle: <surface gone deep on> filed: #a, #b, #c (+1'd: #x, #y) verdict: would this persona adopt today? yes/no - <top blocker, one line> ``` (Exception: `drive-by-critic` runs replace the verdict line with: `verdict: does the front page survive a hostile thread? yes/no - <top-voted dunk, one line>`.) No finding details in this log. Later runs read ONLY this issue before their cold read (to pick an uncovered persona/angle and self-assign a run number); finding details here would contaminate their first impressions. ## Exit condition (the release trust gate) The sweep is done when two consecutive runs produce **zero new adoption-fatal or major-friction findings** and their verdicts are **yes**. At that point the `pressure-test` label is the launch burndown: fix adoption-fatal + major-friction, batch or defer polish, ship. ## Known anchors (pre-existing - comment there instead of re-filing) - ward#414 - add github equivalents for forgejo-paired functionality - ward#208 - ward website - ward#190 - pressure test ward without ward or cli-guard on disk - ward#201 - "blind" openclaw / opencode agent pressure testing - ward#255 - launch writeup (doctrine: every demo shows a denial) - ward#195 - doctor pattern for onboarding onto the ward agent surface - ward#427 - lift the README length cap Filed by Claude Code during the release pressure-test setup session.

coilyco-ops added the

pressure-test

label

2026-07-01 20:24:03 +00:00

coilyco-ops commented

2026-07-01 21:22:32 +00:00

Author

Owner

run: 1
persona: github-only-dev
angle: front-page
filed: none
verdict: would this persona adopt today? yes - the README states the Forgejo dependency and install path up front, so there is no first-screen surprise.

run: 1 persona: github-only-dev angle: front-page filed: none verdict: would this persona adopt today? yes - the README states the Forgejo dependency and install path up front, so there is no first-screen surprise.

coilyco-ops commented

2026-07-01 21:28:52 +00:00

Author

Owner

run: 2
persona: vibe-coder
angle: front-page
filed: none
verdict: would this persona adopt today? yes - the front page gives an honest enough is-this-for-you signal and a direct install path without hiding the Forgejo dependency.

run: 2 persona: vibe-coder angle: front-page filed: none verdict: would this persona adopt today? yes - the front page gives an honest enough is-this-for-you signal and a direct install path without hiding the Forgejo dependency.

coilyco-ops commented

2026-07-01 21:39:05 +00:00

Author

Owner

run: 3
persona: forgejo-selfhoster
angle: yaml-config
filed: #433, #434
verdict: would this persona adopt today? no - the .ward/ward.yaml schema is undocumented anywhere (ward or cli-guard docs), so a self-hoster cannot author a config for their own repo without guessing.

run: 3 persona: forgejo-selfhoster angle: yaml-config filed: #433, #434 verdict: would this persona adopt today? no - the .ward/ward.yaml schema is undocumented anywhere (ward or cli-guard docs), so a self-hoster cannot author a config for their own repo without guessing.

coilyco-ops commented

2026-07-01 21:40:45 +00:00

Author

Owner

run: 3
persona: agent-power-user
angle: install-path
filed: none
verdict: would this persona adopt today? yes - the tap formula, release assets, and README install path line up cleanly, so there is no hidden install wall.

run: 3 persona: agent-power-user angle: install-path filed: none verdict: would this persona adopt today? yes - the tap formula, release assets, and README install path line up cleanly, so there is no hidden install wall.

coilyco-ops commented

2026-07-01 21:40:45 +00:00

Author

Owner

run: 4
persona: security-skeptic
angle: security-posture
filed: none (+1: #430)
verdict: would this persona adopt today? no - SECURITY.md still claims there are no published releases, which is now false and undermines the trust story.

run: 4 persona: security-skeptic angle: security-posture filed: none (+1: #430) verdict: would this persona adopt today? no - SECURITY.md still claims there are no published releases, which is now false and undermines the trust story.

coilyco-ops commented

2026-07-01 21:40:45 +00:00

Author

Owner

run: 6
persona: agent-power-user
angle: agent-legibility
filed: none
verdict: would this persona adopt today? yes - the agent roster and per-harness docs are directly linked, generated where needed, and readable without code spelunking.

run: 6 persona: agent-power-user angle: agent-legibility filed: none verdict: would this persona adopt today? yes - the agent roster and per-harness docs are directly linked, generated where needed, and readable without code spelunking.

coilyco-ops commented

2026-07-01 21:40:45 +00:00

Author

Owner

run: 5
persona: oss-agent-power-user
angle: front-page
filed: none
verdict: would this persona adopt today? yes - the README names claude, goose, codex, and qwen/opencode support up front and the OSS harness pages exist, so the path is visible.

run: 5 persona: oss-agent-power-user angle: front-page filed: none verdict: would this persona adopt today? yes - the README names claude, goose, codex, and qwen/opencode support up front and the OSS harness pages exist, so the path is visible.

coilyco-ops commented

2026-07-01 21:45:48 +00:00

Author

Owner

run: 7
persona: drive-by-critic
angle: release-story
filed: none (+1: #430)
verdict: does the front page survive a hostile thread? no - SECURITY.md says there are no published releases, which the repo contradicts.

run: 7 persona: drive-by-critic angle: release-story filed: none (+1: #430) verdict: does the front page survive a hostile thread? no - SECURITY.md says there are no published releases, which the repo contradicts.

coilyco-ops commented

2026-07-01 21:48:05 +00:00

Author

Owner

run: 2
persona: the-agent-itself
angle: yaml-config
filed: #436, #437, #438, #439 (+1'd: none)
verdict: would this persona adopt today? no - the security: block (the product's core) cannot be authored from docs alone, and a wrong guess passes through silently

run: 2 persona: the-agent-itself angle: yaml-config filed: #436, #437, #438, #439 (+1'd: none) verdict: would this persona adopt today? no - the security: block (the product's core) cannot be authored from docs alone, and a wrong guess passes through silently

coilyco-ops commented

2026-07-01 21:48:59 +00:00

Author

Owner

run: 4
persona: the-agent-itself
angle: kdl-config
filed: #440
verdict: would this persona adopt today? no - drafting a guardfile from ward's docs alone dead-ends in a two-file circular loop (ward-kdl-authoring.md <-> cmd/ward-kdl/README.md); the actual KDL grammar exists only in the unlinked sibling repo cli-guard.

run: 4 persona: the-agent-itself angle: kdl-config filed: #440 verdict: would this persona adopt today? no - drafting a guardfile from ward's docs alone dead-ends in a two-file circular loop (ward-kdl-authoring.md <-> cmd/ward-kdl/README.md); the actual KDL grammar exists only in the unlinked sibling repo cli-guard.

coilyco-ops commented

2026-07-01 21:51:52 +00:00

Author

Owner

test

coilyco-ops commented

2026-07-01 21:51:59 +00:00

Author

Owner

run: 8
persona: agent-legibility
angle: docs-ia
filed: none (+1'd: #429, #433, #438)
verdict: would this persona adopt today? no - the docs point at dead or contradictory public entry points before a newcomer gets a linear path.

run: 8 persona: agent-legibility angle: docs-ia filed: none (+1'd: #429, #433, #438) verdict: would this persona adopt today? no - the docs point at dead or contradictory public entry points before a newcomer gets a linear path.

coilyco-ops commented

2026-07-01 21:55:39 +00:00

Author

Owner

run: 12
persona: forgejo-selfhoster
angle: install-path
filed: #441, #442 (+1'd: none; dedup notes on #433, #436/#434, #437/#440)
verdict: would this persona adopt today? no - the shipped binary is compiled to forgejo.coilysiren.me with a coily* owner gate and an org-private SSM token path; a self-hoster cannot retarget it and no doc says so

run: 12 persona: forgejo-selfhoster angle: install-path filed: #441, #442 (+1'd: none; dedup notes on #433, #436/#434, #437/#440) verdict: would this persona adopt today? no - the shipped binary is compiled to forgejo.coilysiren.me with a coily* owner gate and an org-private SSM token path; a self-hoster cannot retarget it and no doc says so

coilyco-ops commented

2026-07-01 21:56:40 +00:00

Author

Owner

run: 5
persona: drive-by-critic
angle: front-page
filed: #444, #445
verdict: does the front page survive a hostile thread? no - top dunk: "four paragraphs in and I still dont know what problem this solves" (the concrete What it does sentence is buried under naming lore and unexplained bare issue refs).

run: 5 persona: drive-by-critic angle: front-page filed: #444, #445 verdict: does the front page survive a hostile thread? no - top dunk: "four paragraphs in and I still dont know what problem this solves" (the concrete What it does sentence is buried under naming lore and unexplained bare issue refs).

coilyco-ops commented

2026-07-01 21:57:56 +00:00

Author

Owner

run: 15
persona: vibe-coder
angle: failure-paths
filed: #447 (+1'd: #195)
verdict: would this persona adopt today? no - when a warded run fails silently there is no symptom-indexed path from the README to the fix, and this persona never reads subsystem docs

run: 15 persona: vibe-coder angle: failure-paths filed: #447 (+1'd: #195) verdict: would this persona adopt today? no - when a warded run fails silently there is no symptom-indexed path from the README to the fix, and this persona never reads subsystem docs

coilyco-ops commented

2026-07-01 21:59:03 +00:00

Author

Owner

run: 6
persona: oss-agent-power-user
angle: jargon-audit
filed: #448
verdict: would this persona adopt today? no - the READMEs first sentence names qwen as one of four supported harnesses, but qwen is a deprecated alias for opencode with no doc at the expected path, and four different docs disagree on whether it currently works at all.

run: 6 persona: oss-agent-power-user angle: jargon-audit filed: #448 verdict: would this persona adopt today? no - the READMEs first sentence names qwen as one of four supported harnesses, but qwen is a deprecated alias for opencode with no doc at the expected path, and four different docs disagree on whether it currently works at all.

coilyco-ops commented

2026-07-01 21:59:13 +00:00

Author

Owner

run: 17
persona: github-only-dev
angle: jargon-audit
filed: #449 (+1'd: #444, #446)
verdict: would this persona adopt today? no - the README front-loads naming lore and bare issue refs (see #444) before stating what the tool does, and a GitHub arrival cannot resolve any of the refs

run: 17 persona: github-only-dev angle: jargon-audit filed: #449 (+1'd: #444, #446) verdict: would this persona adopt today? no - the README front-loads naming lore and bare issue refs (see #444) before stating what the tool does, and a GitHub arrival cannot resolve any of the refs

coilyco-ops commented

2026-07-01 22:01:52 +00:00

Author

Owner

run: 18
persona: security-skeptic
angle: release-story
filed: none (+1: #430)
verdict: would this persona adopt today? no - SECURITY.md still claims there are no published releases, which the repo contradicts.

run: 18 persona: security-skeptic angle: release-story filed: none (+1: #430) verdict: would this persona adopt today? no - SECURITY.md still claims there are no published releases, which the repo contradicts.

coilyco-ops commented

2026-07-01 22:02:16 +00:00

Author

Owner

run: 7
persona: security-skeptic
angle: failure-paths
filed: #450
verdict: would this persona adopt today? no - ward doctor exits 0 on a repo with zero security: config and neither doctor.md nor SECURITY.md say whether that is deliberate or a gap, so a CI gate built on ward doctor cannot be trusted at face value.

run: 7 persona: security-skeptic angle: failure-paths filed: #450 verdict: would this persona adopt today? no - ward doctor exits 0 on a repo with zero security: config and neither doctor.md nor SECURITY.md say whether that is deliberate or a gap, so a CI gate built on ward doctor cannot be trusted at face value.

coilyco-ops commented

2026-07-01 22:07:03 +00:00

Author

Owner

run: 8
persona: the-agent-itself
angle: install-path
filed: none (+1'd: #442)
verdict: would this persona adopt today? yes - the brew formula compiles from a live, correctly-pinned source tarball; the only issues found (stale ward drive comment, unexplained Linux-only release assets) were already caught by run 12 and dont block install.

run: 8 persona: the-agent-itself angle: install-path filed: none (+1'd: #442) verdict: would this persona adopt today? yes - the brew formula compiles from a live, correctly-pinned source tarball; the only issues found (stale ward drive comment, unexplained Linux-only release assets) were already caught by run 12 and dont block install.

coilyco-ops commented

2026-07-01 22:08:11 +00:00

Author

Owner

run: 21
persona: agent-power-user
angle: kdl-config
filed: none
verdict: would this persona adopt today? no - the guardfile/KDL path still dead-ends in cross-linked docs, and the public surfaces never give a cold reader a complete authoring contract.

run: 21 persona: agent-power-user angle: kdl-config filed: none verdict: would this persona adopt today? no - the guardfile/KDL path still dead-ends in cross-linked docs, and the public surfaces never give a cold reader a complete authoring contract.

coilyco-ops commented

2026-07-01 22:08:22 +00:00

Author

Owner

run: 9
persona: security-skeptic
angle: docs-ia
filed: #451
verdict: would this persona adopt today? no - SECURITY.md never mentions the agent/container/credential surface at all, only the dev-verb gate, so the exact thing this persona is evaluating (letting an agent loose with real credentials) has no stated vulnerability scope or threat model.

run: 9 persona: security-skeptic angle: docs-ia filed: #451 verdict: would this persona adopt today? no - SECURITY.md never mentions the agent/container/credential surface at all, only the dev-verb gate, so the exact thing this persona is evaluating (letting an agent loose with real credentials) has no stated vulnerability scope or threat model.

coilyco-ops commented

2026-07-01 22:10:12 +00:00

Author

Owner

run: 10
persona: drive-by-critic
angle: security-posture
filed: #452
verdict: does the front page survive a hostile thread? no - top dunk: "it will refuse Y, and prove it - except the Claude Code hook, which is best-effort and fails open on a bad YAML file. Which is it?"

run: 10 persona: drive-by-critic angle: security-posture filed: #452 verdict: does the front page survive a hostile thread? no - top dunk: "it will refuse Y, and prove it - except the Claude Code hook, which is best-effort and fails open on a bad YAML file. Which is it?"

coilyco-ops commented

2026-07-01 22:13:41 +00:00

Author

Owner

run: 11
persona: oss-agent-power-user
angle: install-path
filed: #456
verdict: would this persona adopt today? no - agent-goose.md and agent-codex.md say image-baked/no-self-install, container.md says both drop to a shell (not yet imaged); the docs cannot even agree on whether the OSS harnesses launch.

run: 11 persona: oss-agent-power-user angle: install-path filed: #456 verdict: would this persona adopt today? no - agent-goose.md and agent-codex.md say image-baked/no-self-install, container.md says both drop to a shell (not yet imaged); the docs cannot even agree on whether the OSS harnesses launch.

coilyco-ops commented

2026-07-01 22:15:52 +00:00

Author

Owner

run: 12
persona: github-only-dev
angle: install-path
filed: #461
verdict: would this persona adopt today? no - the README presents dev-verb wrapping and Forgejo issue-carrying as one Forgejo-bound pipeline, but exec-verb.md confirms the dev-verb half is entirely forge-agnostic; a GitHub-only reader bounces off a requirement that does not actually apply to half the tool.

run: 12 persona: github-only-dev angle: install-path filed: #461 verdict: would this persona adopt today? no - the README presents dev-verb wrapping and Forgejo issue-carrying as one Forgejo-bound pipeline, but exec-verb.md confirms the dev-verb half is entirely forge-agnostic; a GitHub-only reader bounces off a requirement that does not actually apply to half the tool.

coilyco-ops commented

2026-07-01 22:20:17 +00:00

Author

Owner

run: 13
persona: vibe-coder
angle: jargon-audit
filed: #462
verdict: would this persona adopt today? no - AGENTS.md, the file this personas delegated agent would actually follow, instructs running coily (a tool the same doc calls retiring and that a stranger would not have installed) instead of the native ward ci watch replacement it names two paragraphs earlier.

run: 13 persona: vibe-coder angle: jargon-audit filed: #462 verdict: would this persona adopt today? no - AGENTS.md, the file this personas delegated agent would actually follow, instructs running coily (a tool the same doc calls retiring and that a stranger would not have installed) instead of the native ward ci watch replacement it names two paragraphs earlier.

coilyco-ops commented

2026-07-01 22:23:10 +00:00

Author

Owner

run: 14
persona: forgejo-selfhoster
angle: front-page
filed: none (+1'd: #441)
verdict: would this persona adopt today? no - same root cause as #441: the front page never uses self-hosting language, consistent with the hardcoded org/endpoint that blocks this personas actual adoption.

run: 14 persona: forgejo-selfhoster angle: front-page filed: none (+1'd: #441) verdict: would this persona adopt today? no - same root cause as #441: the front page never uses self-hosting language, consistent with the hardcoded org/endpoint that blocks this personas actual adoption.

coilyco-ops commented

2026-07-01 22:24:36 +00:00

Author

Owner

run: 28
persona: forgejo-selfhoster
angle: agent-legibility
filed: none (+1: #429, #448, #451, #456, #461, #462)
verdict: would this persona adopt today? no - the public docs still leave too many mismatched or dead-end entry points for a cold reader to trust the setup path.

run: 28 persona: forgejo-selfhoster angle: agent-legibility filed: none (+1: #429, #448, #451, #456, #461, #462) verdict: would this persona adopt today? no - the public docs still leave too many mismatched or dead-end entry points for a cold reader to trust the setup path.

coilyco-ops commented

2026-07-01 22:28:04 +00:00

Author

Owner

run: 15
persona: agent-power-user
angle: front-page
filed: #464
verdict: would this persona adopt today? no - Docker is the entire mechanism behind ward agent/warded and is never named as a prerequisite anywhere in README, AGENTS.md, or agent.md; this persona finds out only by hitting a raw error.

run: 15 persona: agent-power-user angle: front-page filed: #464 verdict: would this persona adopt today? no - Docker is the entire mechanism behind ward agent/warded and is never named as a prerequisite anywhere in README, AGENTS.md, or agent.md; this persona finds out only by hitting a raw error.

coilyco-ops commented

2026-07-01 22:33:07 +00:00

Author

Owner

run: 16
persona: the-agent-itself
angle: agent-legibility
filed: #465
verdict: would this persona adopt today? yes, with friction - the docs are technically thorough and accurate but agent-preflight.md in particular buries load-bearing behavior in dense run-on prose, adding context cost for a harness consuming it.

run: 16 persona: the-agent-itself angle: agent-legibility filed: #465 verdict: would this persona adopt today? yes, with friction - the docs are technically thorough and accurate but agent-preflight.md in particular buries load-bearing behavior in dense run-on prose, adding context cost for a harness consuming it.

coilyco-ops commented

2026-07-01 22:34:04 +00:00

Author

Owner

run: 29
persona: agent-power-user
angle: feature-gaps
filed: #467 (+1'd: #464, #441)
verdict: would this persona adopt today? no - no first-run path names prerequisites/auth/dry-run before attempting warded.

run: 29 persona: agent-power-user angle: feature-gaps filed: #467 (+1'd: #464, #441) verdict: would this persona adopt today? no - no first-run path names prerequisites/auth/dry-run before attempting warded.

coilyco-ops commented

2026-07-01 22:36:23 +00:00

Author

Owner

run: 30
persona: vibe-coder
angle: agent-legibility
filed: #468, #469
verdict: would this persona adopt today? no - the front page advertises a retired --watch command, and the agent entrypoint points driver setup at internal source instead of public docs.

run: 30 persona: vibe-coder angle: agent-legibility filed: #468, #469 verdict: would this persona adopt today? no - the front page advertises a retired --watch command, and the agent entrypoint points driver setup at internal source instead of public docs.

coilyco-ops commented

2026-07-01 22:38:15 +00:00

Author

Owner

run: 31
persona: oss-agent-power-user
angle: docs-ia
filed: #470, #471
verdict: would this persona adopt today? no - the local harness path never explains a non-Coily Ollama setup, and goose docs contradict themselves on GO/NO-GO preflight coverage.

run: 31 persona: oss-agent-power-user angle: docs-ia filed: #470, #471 verdict: would this persona adopt today? no - the local harness path never explains a non-Coily Ollama setup, and goose docs contradict themselves on GO/NO-GO preflight coverage.

coilyco-ops commented

2026-07-01 22:39:32 +00:00

Author

Owner

run: 32
persona: security-skeptic
angle: security-posture
filed: #473
verdict: would this persona adopt today? no - the container safety proof claims a one-repo blast radius even though documented --repo grants additional writable repos.

run: 32 persona: security-skeptic angle: security-posture filed: #473 verdict: would this persona adopt today? no - the container safety proof claims a one-repo blast radius even though documented --repo grants additional writable repos.

coilyco-ops commented

2026-07-01 22:42:08 +00:00

Author

Owner

run: 30
persona: agent-power-user
angle: feature-gaps
filed: #474, #475 (+1'd: #414)
verdict: would this persona adopt today? no - three of five day-1 must-haves (forge-free path, stated prerequisites, testable guard policy) are absent from docs alone, and the README's own watch-the-run instruction errors as a retired flag.

run: 30 persona: agent-power-user angle: feature-gaps filed: #474, #475 (+1'd: #414) verdict: would this persona adopt today? no - three of five day-1 must-haves (forge-free path, stated prerequisites, testable guard policy) are absent from docs alone, and the README's own watch-the-run instruction errors as a retired flag.

coilyco-ops commented

2026-07-01 22:43:18 +00:00

Author

Owner

run: 31
persona: vibe-coder
angle: agent-legibility
filed: #476 (+1'd: #462, #468; note on #436)
verdict: would this persona adopt today? no - their delegated agent's entry point (AGENTS.md) opens with a private-workspace reference it cannot resolve and never addresses the set-this-up-for-me task, so the delegation dead-ends into guesswork.

run: 31 persona: vibe-coder angle: agent-legibility filed: #476 (+1'd: #462, #468; note on #436) verdict: would this persona adopt today? no - their delegated agent's entry point (AGENTS.md) opens with a private-workspace reference it cannot resolve and never addresses the set-this-up-for-me task, so the delegation dead-ends into guesswork.

coilyco-ops commented

2026-07-01 22:48:11 +00:00

Author

Owner

run: 32
persona: github-only-dev
angle: release-story
filed: #477 (+1'd: #438; baseline comment on #454)
verdict: would this persona adopt today? no - the GitHub surface this persona actually reads shows releases frozen at v0.5.8 under a v0.24x README, so the release story affirmatively misstates whether the project is alive.

run: 32 persona: github-only-dev angle: release-story filed: #477 (+1'd: #438; baseline comment on #454) verdict: would this persona adopt today? no - the GitHub surface this persona actually reads shows releases frozen at v0.5.8 under a v0.24x README, so the release story affirmatively misstates whether the project is alive.

coilyco-ops commented

2026-07-01 22:51:33 +00:00

Author

Owner

run: 33
persona: oss-agent-power-user
angle: docs-ia
filed: #481
verdict: would this persona adopt today? no - LICENSE and per-harness pages pass, but the documented contribution process dead-ends at a disabled Forgejo sign-up and a read-only mirror, so investing in the ecosystem is structurally impossible.

run: 33 persona: oss-agent-power-user angle: docs-ia filed: #481 verdict: would this persona adopt today? no - LICENSE and per-harness pages pass, but the documented contribution process dead-ends at a disabled Forgejo sign-up and a read-only mirror, so investing in the ecosystem is structurally impossible.

coilyco-ops commented

2026-07-01 22:53:58 +00:00

Author

Owner

run: 17
persona: the-agent-itself
angle: first-run
filed: #478, #480 (+1'd: #441, #467)
verdict: would this persona adopt today? no - a fresh brew install cleanly succeeds, but the documented README Usage block breaks on line 3 (ward lint does not exist in the shipped binary), and the Makefile-target validation ward doctor runs instead false-positives on a completely standard Makefile.

run: 17 persona: the-agent-itself angle: first-run filed: #478, #480 (+1'd: #441, #467) verdict: would this persona adopt today? no - a fresh brew install cleanly succeeds, but the documented README Usage block breaks on line 3 (ward lint does not exist in the shipped binary), and the Makefile-target validation ward doctor runs instead false-positives on a completely standard Makefile.

coilyco-ops commented

2026-07-01 22:54:42 +00:00

Author

Owner

run: 34
persona: security-skeptic
angle: front-page
filed: #482, #483 (+1'd: #430, #451)
verdict: would this persona adopt today? no - the front page defines the product by a deny-list the project's own container docs say does not exist anywhere, so the stated guarantee and the honest internal docs cannot both be true.

run: 34 persona: security-skeptic angle: front-page filed: #482, #483 (+1'd: #430, #451) verdict: would this persona adopt today? no - the front page defines the product by a deny-list the project's own container docs say does not exist anywhere, so the stated guarantee and the honest internal docs cannot both be true.

coilyco-ops commented

2026-07-01 22:56:36 +00:00

Author

Owner

run: 18
persona: drive-by-critic
angle: install-path
filed: none (comment on planned-direction anchor #454)
verdict: does the front page survive a hostile thread? no - top dunk: "3 minutes and a Go toolchain to install a CLI tool - wheres the binary?" Already tracked by planned direction #454 (full binary matrix), so commented there with real timing data instead of filing fresh.

run: 18 persona: drive-by-critic angle: install-path filed: none (comment on planned-direction anchor #454) verdict: does the front page survive a hostile thread? no - top dunk: "3 minutes and a Go toolchain to install a CLI tool - wheres the binary?" Already tracked by planned direction #454 (full binary matrix), so commented there with real timing data instead of filing fresh.

coilyco-ops commented

2026-07-01 22:58:07 +00:00

Author

Owner

run: 19
persona: security-skeptic
angle: yaml-config
filed: none (+1'd: #436)
verdict: would this persona adopt today? no - same root cause as #436: zero worked security: block examples anywhere in the repo, so the products actual value proposition cannot be verified from docs alone.

run: 19 persona: security-skeptic angle: yaml-config filed: none (+1'd: #436) verdict: would this persona adopt today? no - same root cause as #436: zero worked security: block examples anywhere in the repo, so the products actual value proposition cannot be verified from docs alone.

coilyco-ops commented

2026-07-01 22:58:09 +00:00

Author

Owner

run: 35
persona: forgejo-selfhoster
angle: failure-paths
filed: #484 (+1'd: #450; confirm comment on #195)
verdict: would this persona adopt today? no - their first inevitable failure (trust-gate refusal on their own org) has no documented explanation, no doctor coverage, and no next step anywhere in the docs.

run: 35 persona: forgejo-selfhoster angle: failure-paths filed: #484 (+1'd: #450; confirm comment on #195) verdict: would this persona adopt today? no - their first inevitable failure (trust-gate refusal on their own org) has no documented explanation, no doctor coverage, and no next step anywhere in the docs.

coilyco-ops commented

2026-07-01 23:00:01 +00:00

Author

Owner

run: 36
persona: drive-by-critic
angle: docs-ia
filed: none (+1'd: #446; substance comments on #460, #453)
verdict: does the front page survive a hostile thread? no - top-voted dunk: 'the README's own commands don't exist in the shipped binary' (ward lint per #478, --watch per #468/#474), with 'docker plus a YAML allowlist, and 23 reference pages for the maintainer's trello guardfile' close behind.

run: 36 persona: drive-by-critic angle: docs-ia filed: none (+1'd: #446; substance comments on #460, #453) verdict: does the front page survive a hostile thread? no - top-voted dunk: 'the README's own commands don't exist in the shipped binary' (ward lint per #478, --watch per #468/#474), with 'docker plus a YAML allowlist, and 23 reference pages for the maintainer's trello guardfile' close behind.

coilyco-ops commented

2026-07-01 23:00:05 +00:00

Author

Owner

run: 20
persona: agent-power-user
angle: security-posture
filed: none (+1'd: #482)
verdict: would this persona adopt today? no - same root cause as #482/#451: the credential/container security story is honestly documented in isolation but contradicts the READMEs deny-list framing, and codex specifically runs danger-full-access with zero approval gating inside the container.

run: 20 persona: agent-power-user angle: security-posture filed: none (+1'd: #482) verdict: would this persona adopt today? no - same root cause as #482/#451: the credential/container security story is honestly documented in isolation but contradicts the READMEs deny-list framing, and codex specifically runs danger-full-access with zero approval gating inside the container.

coilyco-ops commented

2026-07-01 23:00:48 +00:00

Author

Owner

run: 21
persona: the-agent-itself
angle: jargon-audit
filed: none (+1'd: #478)
verdict: would this persona adopt today? yes for CONTRIBUTING.md itself (clean, well-scoped, low jargon) - but it independently reconfirms #478 (ward lint documented, does not exist), now hit in 3 separate docs.

run: 21 persona: the-agent-itself angle: jargon-audit filed: none (+1'd: #478) verdict: would this persona adopt today? yes for CONTRIBUTING.md itself (clean, well-scoped, low jargon) - but it independently reconfirms #478 (ward lint documented, does not exist), now hit in 3 separate docs.

coilyco-ops commented

2026-07-01 23:01:22 +00:00

Author

Owner

run: 37
persona: the-agent-itself
angle: failure-paths
filed: #485 (comment on #447)
verdict: would this persona adopt today? no - a supervising agent cannot tell from public docs whether a dispatched run launched, was refused, or failed; the exit-code and outcome contract exists only in Go.

run: 37 persona: the-agent-itself angle: failure-paths filed: #485 (comment on #447) verdict: would this persona adopt today? no - a supervising agent cannot tell from public docs whether a dispatched run launched, was refused, or failed; the exit-code and outcome contract exists only in Go.

coilyco-ops commented

2026-07-01 23:02:09 +00:00

Author

Owner

run: 22
persona: forgejo-selfhoster
angle: release-story
filed: #486
verdict: would this persona adopt today? no - release notes are raw commit-message dumps of internal refactor jargon with no user-facing framing, so this persona cannot tell whether any given release (or upgrade) affects them without reading source history.

run: 22 persona: forgejo-selfhoster angle: release-story filed: #486 verdict: would this persona adopt today? no - release notes are raw commit-message dumps of internal refactor jargon with no user-facing framing, so this persona cannot tell whether any given release (or upgrade) affects them without reading source history.

coilyco-ops commented

2026-07-01 23:09:28 +00:00

Author

Owner

run: 23
persona: oss-agent-power-user
angle: failure-paths
filed: #487
verdict: would this persona adopt today? no - claude gets a documented anti-hang credential smoke test; goose and opencode explicitly opt out of any smoke gate with zero substitute failure-path docs for an unreachable Ollama endpoint, the exact asymmetry this persona is built to catch.

run: 23 persona: oss-agent-power-user angle: failure-paths filed: #487 verdict: would this persona adopt today? no - claude gets a documented anti-hang credential smoke test; goose and opencode explicitly opt out of any smoke gate with zero substitute failure-path docs for an unreachable Ollama endpoint, the exact asymmetry this persona is built to catch.

coilyco-ops commented

2026-07-01 23:11:21 +00:00

Author

Owner

run: 24
persona: the-agent-itself
angle: docs-ia
filed: #488
verdict: would this persona adopt today? no - docs/ has 92 files and zero index/landing page; the only link that could have served as one (AGENTS.md docs/README.md, #429) is dead, so there is no way to browse into the doc set, only to follow outbound links from README/FEATURES.md/AGENTS.md.

run: 24 persona: the-agent-itself angle: docs-ia filed: #488 verdict: would this persona adopt today? no - docs/ has 92 files and zero index/landing page; the only link that could have served as one (AGENTS.md docs/README.md, #429) is dead, so there is no way to browse into the doc set, only to follow outbound links from README/FEATURES.md/AGENTS.md.

coilyco-ops commented

2026-07-01 23:12:15 +00:00

Author

Owner

run: 25
persona: github-only-dev
angle: failure-paths
filed: none
verdict: would this persona adopt today? yes for this angle specifically - agent-wrong-repo.md is clean and self-contained (ties into the already-filed #484 trust-set gap, not a new issue), and LICENSE/CODE_OF_CONDUCT.md are both standard, complete, no template placeholders left unfilled.

run: 25 persona: github-only-dev angle: failure-paths filed: none verdict: would this persona adopt today? yes for this angle specifically - agent-wrong-repo.md is clean and self-contained (ties into the already-filed #484 trust-set gap, not a new issue), and LICENSE/CODE_OF_CONDUCT.md are both standard, complete, no template placeholders left unfilled.

coilyco-ops added this to the ward launch milestone

2026-07-01 23:29:44 +00:00

coilyco-ops added the

labels

2026-07-01 23:30:45 +00:00

coilyco-ops commented

2026-07-01 23:59:31 +00:00

Author

Owner

Triage note: this sits in the milestone as the sweep tracker, not dispatchable work - it closes when the exit condition is met (two consecutive clean runs, both verdicts yes). Engineers should not pick this up. Recorded by Claude Code (Fable) during the 2026-07-01 ward launch triage session with Kai.

coilyco-ops referenced this issue

2026-07-02 00:01:42 +00:00

meta: jul 1st evening sprint - launch burndown session 1 #492

coilyco-ops referenced this issue

2026-07-02 00:18:00 +00:00

Pressure test ward without ward or CLI guard on disk #190

coilyco-ops referenced this issue

2026-07-02 00:21:39 +00:00

"blind" openclaw / opencode agent pressure testing #201

coilysiren referenced this issue

2026-07-02 07:10:55 +00:00

meta: jul 1st evening sprint - launch burndown session 1 #492

coilyco-ops added

and removed

labels

2026-07-10 09:00:44 +00:00

coilyco-ops added

and removed

labels

2026-07-29 04:38:15 +00:00

coilyco-ops commented

2026-07-31 07:30:16 +00:00

Author

Owner

Closing this historical coordination artifact. The launch milestone is closed, the run log remains available for reference, and #1610 plus #1620 to #1623 now define the current sealing boundary.

coilyco-ops closed this issue

2026-07-31 07:30:18 +00:00

Rows
Columns

meta: release pressure-test run log (cold-read usability sweeps) #428

What this is

Protocol for each run

Exit condition (the release trust gate)

Known anchors (pre-existing - comment there instead of re-filing)