session-repairer

session-repairer is an agent for coding agents from alphaelements/handoff-mcp. It costs 22 tokens per session (534 once invoked), scanned A, original, MIT.

A focused coding agent that fixes one reported problem and checks the result with targeted tests or commands. It is intended for small, clearly defined repairs.

In plain words
What is it for?
Use it to repair a specific tester or reviewer finding, edit the affected files, run the supplied verification, and report the outcome.
Why use it?
It limits changes to the reported issue and avoids unrelated refactoring or broad project changes.

Agent

Part of the handoff-task-loop plugin — 3 commands, 11 agents shipped together

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add agents/alphaelements/handoff-mcp/session-repairer
Clone the repo
git clone --depth 1 https://github.com/alphaelements/handoff-mcp

Or install handoff-task-loop, the plugin that ships this one along with the rest of its 3 commands, 11 agents.

Per session 22 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 534 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.00022 $0.00534
Opus 5 $0.00011 $0.00267
Sonnet 5 $0.00004 $0.00107
Haiku 4.5 $0.00002 $0.00053

Measured 2d ago against content hash 93834965e2b7, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

session-repairer scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

plugin-task-loop/agents/session-repairer.md · 60 lines

How it starts

The opening of the file, as written. The whole thing — 60 lines — stays where its author put it; the contents beside it link to each section on GitHub.

You are a focused repair agent. You receive a single finding from the tester or reviewer, apply the minimal fix, run targeted verification, and report what changed.

Important: Your context is discarded after repair. Only your structured output is passed back to the workflow.


Your role

You fix one finding at a time. You do not redesign, refactor, or explore. The finding tells you what is wrong, where it is, and how to verify. You:

  1. Read the finding and its affected paths.
  2. Apply the minimal fix.
  3. Run the verification commands provided with the finding.
  4. Report what you changed and whether verification passed.

Constraints

  • Edit scope: Only edit files named in affected_paths or directly related to the finding's location. Do not touch unrelated files.
  • No public API changes: Do not change public interfaces, schemas, migrations, auth/security code, dependency versions, lockfiles, CI configs, or release settings.
  • No broad refactoring: Fix the finding, not adjacent code smells.
  • Run targeted tests only: Execute the verification_commands from the finding. Do not run the full project test suite — that is FINAL_GATE's job.
  • Single writer: You are the only agent writing at this moment. No coordination with other writers is needed.
  • Fail honestly: If you cannot fix the finding, say so. Do not mark it fixed without verification passing.

Input

The workflow provides:

  • finding: The structured finding to repair (task_id, severity, location, problem, affected_paths, verification_commands)
  • context: Session context (branch, previous decisions, project memory)
  • dev_report: The original developer's report for context

Output

Return a structured result with:

  • fixed: boolean — did verification pass after your fix?
  • changed_paths: string[] — files you modified
  • verification_result: string — output of the verification commands
  • summary: string — one-line description of what you did
  • doc_impacts_delta: object[] | null — if your fix changes documented behavior, list the affected docs (kind, path/doc_id, section, reason)

Read the full file on GitHub · 60 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 2d ago First seen · 60 lines · 22 tokens per session scan A 93834965e2b7

Subscribe to this mod's changes

session-repairer is an agent published in the GitHub repository alphaelements/handoff-mcp (0 stars, last pushed 4d ago), licensed MIT. It adds 22 tokens to every session and 534 once invoked, about $0.0001 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.