lindigo-devdigest: plugin catalog

What’s new

2026-11-11
  1. architecture-review

    v1.1.0 ·

    Eval suite evals/: architecture-reviewer starts on a pre-merge layering review and not on an unrelated request (deterministic tool_used graders; run with claude plugin eval).

  2. engineering-paved-path

    v1.1.0 ·

    Eval suite evals/: onion-architecture loads on a backend layering question, frontend-ui-architecture on a component-structure question, neither on an unrelated request (deterministic tool_used graders; run with claude plugin eval).

  3. research-tools

    v1.1.0 ·

    Eval suite evals/: researcher starts on a research question that needs evidence and not on a trivial edit (deterministic tool_used graders; run with claude plugin eval).

  4. sdd-engineering

    v1.1.0 ·

    workflow-retro ships scripts/transcript-metrics.mjs, called through ${CLAUDESKILLDIR} so it works wherever the plugin is installed. It reads the session and subagent transcripts and returns tokens (counted once per message.id), cache hit rate, tool calls, spans, parallelism, the critical path, files read by several agents and, with --prices, cost. It never prints transcript text. Self-test: scripts/transcript-metrics.self-test.mjs.Eval suite evals/: spec-creator starts on a spec request, workflow-retro and fastify-best-practices load on their requests, run-plan never starts on its own (it has disable-model-invocation), and an unrelated request starts no skill or agent of this plugin.

  5. research-tools

    v1.0.2 ·

    Declare the always-on context cost in COMPATIBILITY.md; CI checks it against claude plugin details.

  6. architecture-review

    v1.0.1 ·

    Declare the always-on context cost in COMPATIBILITY.md; CI checks it against claude plugin details.

  7. engineering-paved-path

    v1.0.1 ·

    Declare the always-on context cost in COMPATIBILITY.md; CI checks it against claude plugin details.

  8. research-tools

    v1.0.1 ·

    README: one-command install (claude plugin install research-tools --marketplace …), how to check the install, and how to update or turn on auto-update.

  9. sdd-engineering

    v1.0.1 ·

    Declare the always-on context cost in COMPATIBILITY.md; CI checks it against claude plugin details.

  10. architecture-review

    v1.0.0 ·

    Agent architecture-reviewer: read-only review of a change set against the onion and frontend UI architecture skills and the project's own documented rules; learns the repository layout first; generalized to any repository.Hook bash-guard (PreToolUse on Bash): read-only git only while architecture-reviewer runs; generic script with one read-only profile, agent-to-profile map passed from hooks.json; self-test.

Plugins

4 entries, in dependency order
Plugin

research-tools

research · v1.1.0

Read-only researcher agent: answers one question from your repository or external docs, with evidence for every claim.

Used by sdd-engineering

/plugin install research-tools --marketplace lindigo-design/devdigest-marketplace
Plugin

architecture-review

code-review · v1.1.0

Read-only architecture-reviewer agent that checks a change set against layering rules, with a read-only Bash guard.

Depends on engineering-paved-path ^1.0Used by sdd-engineering

/plugin install architecture-review --marketplace lindigo-design/devdigest-marketplace
Plugin

sdd-engineering

workflow · v1.1.0

Spec-driven workflow: spec-creator, implementation-planner, implementer and plan-verifier agents, run-plan and workflow-retro skills.

Depends on engineering-paved-path ^1.0, research-tools ^1.0, architecture-review ^1.0

/plugin install sdd-engineering --marketplace lindigo-design/devdigest-marketplace