@plainconceptsplatform/agent-harness 2.7.0 → 3.0.1

package/harness/.agents/skills/pc-guardrails-generic/SKILL.md

@@ -1,47 +0,0 @@
----
-name: pc-guardrails-generic
-description: Generic guardrails, foundational rules that all agents follow. Users add specialized guardrails skills for specific concerns. Covers secrets, code quality, security, tool usage, and engineer workflow.
-license: MIT
----
-
-## Transitive loads (optimization skills)
-
-The marker sections below name the optimization skills this project selected. Load each one before doing any work.
-
-## Secrets
-
-- Treat `.env` files as write-only: write to them when configuring, and read credentials at runtime from the environment or the secret store.
-- Never put a credential, API key or token in a log line, an output, or a commit in anything but encrypted or template form. Anything printed is in a CI log that outlives the run.
-
-## Code
-
-- Comments are for WHY, not WHAT. Use them only where the code does something non-obvious or the reason cannot be inferred from context. Past a 10% comment ratio in a file, refactor for clarity instead.
-- Never add a file that collects unrelated things — `constants.js`, `types.ts`, `config.js`, `utils.ts`. One responsibility per file, split by domain or feature (`user-constants.ts`, `order-types.ts`, `auth-config.ts`). A file importing from many unrelated modules is already the symptom.
-
-## Temporary files
-
-- Never write outside `$REPO_ROOT`, and never to an operating-system temporary directory: the next step and the next agent cannot see it, and nobody cleans it up. Scratch goes under `$REPO_ROOT/.opencode/.tmp/`, in a task-specific child directory when needed (enforced by pc-system-reminders).
-- Never report a path under `.tmp/` as a deliverable. Copy or move the artifact to its required repository path first.
-- Never leave scratch files behind at the end of a task, unless they are the evidence for a failure you are reporting.
-
-<!-- PC-GUARDRAILS-RTK-START -->
-<!-- PC-GUARDRAILS-RTK-END -->
-
-<!-- PC-GUARDRAILS-CODEGRAPH-START -->
-<!-- PC-GUARDRAILS-CODEGRAPH-END -->
-
-<!-- PC-GUARDRAILS-MEMORY-START -->
-<!-- PC-GUARDRAILS-MEMORY-END -->
-
-<!-- PC-GUARDRAILS-SIMPLE-ENGLISH-START -->
-<!-- PC-GUARDRAILS-SIMPLE-ENGLISH-END -->
-
-<!-- PC-GUARDRAILS-HUMANIZER-START -->
-<!-- PC-GUARDRAILS-HUMANIZER-END -->
-
-## Engineer workflow (when spawned)
-
-The lead put your task IDs and their text in your prompt. Two things about that are not up to you:
-
-- Load every skill under your `## Abilities` before you start, guardrails first, one `skill` call per `@skill-name`. Editing, shell and spawning are blocked until you have (pc-system-reminders).
-- Edit only files in your assigned scope, then return a summary: task IDs done, files changed, tests and lint result, decisions made. Then you exit. Never poll for more work, and never claim a task the lead did not give you: the lead spawns with the work in hand, so a worker that waits is a worker that hangs the wave.

package/harness/.agents/skills/pc-make-architecture/structure-template.md

@@ -1,38 +0,0 @@
-# ARCHITECTURE.md structure template
-
-Write (or update) `ARCHITECTURE.md` following this structure. Only include sections relevant to the project; omit sections with no evidence.
-
-- Architecture Overview: what the system is, what problem it solves, major architectural style
-- 1. Project Structure: annotated directory tree with purpose of each major directory
-- 2. High-Level System Diagram: Mermaid diagram of actors, services, data stores, external systems
-- 3. Core Components: each major component: name, responsibility, key files, technologies, inputs/outputs
-  - 3.1 Frontend / User Interface (if present)
-  - 3.2 Backend / Server / API (if present)
-  - 3.3 Shared Libraries / Common Code (if present)
-  - 3.4 CLI / Scripts / Automation (if present)
-- 4. Data Flow: request lifecycle, key user journeys, sequence diagram for main runtime flow
-- 5. Data Stores: all persistent storage: type, purpose, schemas, migration approach
-- 6. External Integrations / APIs: each integration: method, config location, auth, failure behavior
-- 7. Key Technologies: full stack summary with architectural relevance of each
-- 8. Deployment & Infrastructure: build artifacts, env config, containerization, CI/CD, hosting
-- 9. Security Architecture: auth, authz, secrets, input validation, trust boundaries
-- 10. Monitoring & Observability: logging, metrics, tracing, error reporting
-- 11. Performance & Scalability: caching, batching, concurrency, known bottlenecks
-- 12. Development Workflow: local setup, install/dev/test/build/lint commands
-- 13. Testing Strategy: test frameworks, locations, coverage gates, gaps
-- 14. Architectural Decisions & Rationale: key choices with evidence and tradeoffs
-- 15. Constraints, Risks, and Technical Debt: tight coupling, TODOs, operational risks
-- 16. Future Considerations: documented roadmap + reasonable recommendations (labeled as such)
-- 17. Project Identification: name, language, type, runtime, date of review, maintainer
-- 18. Glossary / Acronyms: project-specific terms an agent or new developer needs to know
-
-Append at the very end of the file:
-```
-<!-- Last updated: <current ISO timestamp> -->
-```
-
-Rules:
-- Be specific and concrete: include actual directories, files, modules, commands.
-- Mark anything undiscoverable as "Not evident from the repository".
-- Use Mermaid diagrams where helpful.
-- Write as if this document will be committed and maintained over time.

package/harness/.agents/skills/pc-make-evidence-scaffold/SKILL.md

@@ -1,18 +0,0 @@
----
-name: pc-make-evidence-scaffold
-description: DEPRECATED. Visual evidence is now built into pc-ops-evidence using playwright-cli + pnpm run dev. No per-project scaffold is needed. This skill is kept for backward compatibility but should not be used.
-license: MIT
----
-
-# DEPRECATED
-
-This skill is no longer needed. Visual evidence uses a two-phase architecture:
-
-1. **Agent phase:** `pc-ops-evidence` writes a `capturePlan` in `evidence.json` (the agent sandbox cannot run Docker or headless Chromium)
-2. **CI phase:** A separate "Visual evidence" CI workflow reads the capturePlan and captures screenshots on a runner with full Docker and Chrome access
-
-No per-project scaffold, fixture apps, or scenario registries are required. The `pc-ops-evidence` skill handles everything generically.
-
-If you previously ran `/make-evidence-scaffold` and have a `src/visual-evidence/` directory or `visual-evidence` scripts in `package.json`, you can delete them — the new system does not use them.
-
-To capture evidence for a change, run `/ops-evidence`.

package/harness/.agents/skills/pc-make-evidence-scaffold/evidence-contract.md

@@ -1,29 +0,0 @@
-# Evidence contract
-
-Evidence captured by `/ops-evidence` lives at `openspec/changes/archive/<dated>-<id>/evidence/`. A standalone pre-archive capture may use `openspec/changes/<id>/evidence/`, but archive moves it into the archived change before publication. That folder contains only:
-- ordered capture images (`01-{label}.png/webp`, ...) and/or `flow.gif`
-- `evidence.json`: the manifest, schema below.
-
-## version 1 schema
-
-```jsonc
-{
-  "version": 1,
-  "changeId": "...",
-  "required": true,
-  "status": "passed",            // passed | skipped | failed | blocked
-  "assets": [ { "type": "screenshot", "path": "openspec/changes/archive/<dated>-<id>/evidence/01-final.png", "caption": "...", "bytes": 0, "format": "png" } ],
-  "reason": "...",                 // skipped | blocked
-  "failedStep": "...",             // failed
-  "prMarkdown": "## Evidence ..."
-}
-```
-
-## Statuses
-
-- `passed`: evidence required and produced.
-- `skipped`: evidence not required (see decision rule). Exit success.
-- `blocked`: required but could not run (no harness, app won't start, budget exceeded). Not a skip. Surface it.
-- `failed`: a project harness ran and its assertions failed. Surface it.
-
-`blocked` (required but unrunnable) is never treated as a skip.

package/harness/.opencode/commands/make-evidence-scaffold.md

@@ -1,5 +0,0 @@
----
-description: DEPRECATED. Per-project evidence scaffolds are gone; run /ops-evidence instead.
----
-
-Load the `pc-make-evidence-scaffold` skill.

package/harness/.opencode/tui/pc-subagents.tsx

@@ -1,98 +0,0 @@
-/** @jsxImportSource @opentui/solid */
-import type { TuiPlugin, TuiPluginModule } from "@opencode-ai/plugin/tui"
-import { createSignal, For, Show } from "solid-js"
-import { readFile } from "node:fs/promises"
-import { join } from "node:path"
-
-const id = "ob.subagents"
-
-type Row = { id: string; agent: string; model: string; task: string; status: string }
-
-// Renders a live "Subagents" panel in the session sidebar.
-//
-// Server plugins (opencode.json) cannot draw UI, only TUI plugins (tui.json)
-// can, via the `sidebar_content` slot. This panel is fed by the
-// `.opencode/harness-run.json` state file that the `pc-subagent-monitor` server
-// plugin maintains, so the two cooperate: server plugin = data producer,
-// this TUI plugin = renderer.
-const tui: TuiPlugin = async (api) => {
-  const statePath = join(process.cwd(), ".opencode", "harness-run.json")
-  const [rows, setRows] = createSignal<Row[]>([])
-  // Only show live subagents, finished/failed ones are kept in harness-run.json
-  // for recovery but are not navigable targets the user cares about here.
-  // Entries marked stale (left "running" by a crashed process) are not live.
-  const active = () => rows().filter((r) => r.status === "running")
-
-  const refresh = async () => {
-    try {
-      const data = JSON.parse(await readFile(statePath, "utf-8"))
-      const agents = data?.agents ?? {}
-      setRows(
-        Object.entries(agents)
-          .filter(([, a]: [string, any]) => !a?.stale)
-          .map(([sid, a]: [string, any]) => ({
-            id: sid,
-            agent: a?.agent ?? "?",
-            model: a?.model ?? "",
-            task: Array.isArray(a?.tasks) && a.tasks.length ? a.tasks.join(",") : (a?.title ?? ""),
-            status: a?.status ?? "running",
-          })),
-      )
-    } catch (err: any) {
-      // Missing file = genuinely idle. Anything else (transient read/parse
-      // hiccup) keeps the last good rows instead of flashing "idle" mid-run.
-      if (err?.code === "ENOENT") setRows([])
-    }
-  }
-
-  await refresh()
-  // session.updated fires constantly; debounce so we don't re-read the file
-  // on every keystroke-sized event.
-  let timer: ReturnType<typeof setTimeout> | null = null
-  const scheduleRefresh = () => {
-    if (timer) return
-    timer = setTimeout(() => {
-      timer = null
-      void refresh()
-    }, 200)
-  }
-  for (const evt of ["session.created", "session.idle", "session.updated"]) {
-    try {
-      api.event.on(evt as any, scheduleRefresh)
-    } catch {
-      /* event type unavailable on this host, ignore */
-    }
-  }
-
-  api.slots.register({
-    order: 50,
-    slots: {
-      sidebar_content() {
-        return (
-          <box flexDirection="column">
-            <text>Subagents</text>
-            <Show when={active().length === 0}>
-              <text>  idle</text>
-            </Show>
-            <For each={active()}>
-              {(r) => (
-                <box onMouseUp={() => api.route.navigate("session", { sessionID: r.id })}>
-                  <text>
-                    {"▶ "}
-                    {r.agent}
-                    {r.model ? ` · ${r.model}` : ""}
-                    {r.task ? `, ${r.task}` : ""}
-                  </text>
-                </box>
-              )}
-            </For>
-          </box>
-        )
-      },
-    },
-  })
-}
-
-const pluginModule: TuiPluginModule & { id: string } = { id, tui }
-
-export default pluginModule

package/harness/.opencode/tui.json

@@ -1,6 +0,0 @@
-{
-  "$schema": "https://opencode.ai/tui.json",
-  "plugin": [
-    "./tui/pc-subagents.tsx"
-  ]
-}