SerenecoTech

SerenecoTech/agent-skills

Plugin Claude Code

Plugin marketplace listing 3 plugins: adversarial-review, handoff, guard-hooks.

SerenecoTech/agent-skills

Plugin Claude Code

Adversarial code, architecture and specification review using OpenAI Codex as a second hostile reviewer. Two front doors (review-code, review-design) over one shared engine, plus PreToolUse hooks that enforce what a skill cannot: write-path authorisation, refutation-appendix ingestion, and rubric integrity.

SerenecoTech/agent-skills

Hook

Runs before the agent uses a tool for Edit, Write, MultiEdit, NotebookEdit and Bash tool calls, executing write-path-gate.sh, appendix-gate.sh and rubric-gate.sh via bash (3 commands). From SerenecoTech/agent-skills.

SerenecoTech/agent-skills

Hook

Runs after a tool call finishes for Bash tool calls, executing verification-recorder.sh via bash. From SerenecoTech/agent-skills.

SerenecoTech/agent-skills

Skill Claude CodeCodex

Adversarially review code changes for bugs, security issues and correctness, using OpenAI Codex as a second, hostile reviewer. Trigger when the user asks for a code review, wants code checked for bugs or security issues, asks to review a PR, branch, commit, diff or working tree, or says things like "any issues with…

SerenecoTech/agent-skills

Skill Claude CodeCodex

Adversarially review an idea, architecture, design decision, spec or plan, using OpenAI Codex as a second, hostile reviewer. Trigger when the user asks to critique, evaluate, sanity-check, stress-test or debate any concept, design doc, RFC, proposal, specification, plan or requirements — even without mentioning Codex…

SerenecoTech/agent-skills

Plugin Claude Code

Security guardrails enforced by hooks: deny dangerous shell commands and privilege escalation, block writes to credential files and system directories, refuse to read a file that holds a live credential, refuse to print a credential-bearing environment variable, detect secrets in written content, pre-approve…

SerenecoTech/agent-skills

Plugin Claude Code

Session continuity across context limits. The handoff skill writes a HANDOFF.md capturing goal, done/not-done, failed approaches and decisions with their rationale; handoff-resume reads it, drift-checks the repo, rehydrates context and continues without the user re-explaining the task.

SerenecoTech/agent-skills

Skill Claude CodeCodex

Use when a fresh session must pick up in-flight work left by a prior session, such as when the user says resume, continue, or pick up where we left off, or when a HANDOFF.md is present and the task is unfinished.

SerenecoTech/agent-skills

Skill Claude CodeCodex

Use when context is running low, before /clear, the conversation is being compacted, or the user asks to summarize work so a fresh agent can continue a task in a future session.