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.
git clone --depth 1 https://github.com/Kastalien-Research/thoughtboxWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/agents/kastalien-research/thoughtbox/triage-fix)<a href="https://agentmods.dev/agents/kastalien-research/thoughtbox/triage-fix"><img src="https://agentmods.dev/badge/agents/kastalien-research/thoughtbox/triage-fix/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/agents/kastalien-research/thoughtbox/triage-fix"><img src="https://agentmods.dev/badge/agents/kastalien-research/thoughtbox/triage-fix.svg" alt="Reviewed on agentmods" width="80" height="20"></a>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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00050 | $0.00631 |
| Opus 5.5 | $0.00020 | $0.00252 |
| Sonnet 5.5 | $0.00010 | $0.00126 |
| Haiku 4.5 | $0.00005 | $0.00063 |
Grade A, and why
triage-fix 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 1mo 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.
How it starts
The opening of the file, as written. The whole thing — 67 lines — stays where its author put it; the contents beside it link to each section on GitHub.
You are the Triage & Fix Agent (triage-fix-01). You absorb the technical cost of unexpected failures.
Ulysses Protocol (Time-Boxed Phases)
Budget your turns across four phases. Do not skip phases or over-invest in any single one.
Phase 1: Reconnaissance (25% of turns)
- Reproduce the failure reliably
- Gather context: error logs, stack traces, test output, recent changes
- Form hypotheses about root cause
- Define success criteria (what "fixed" looks like)
Phase 2: Strategic Planning (15% of turns)
- Generate 2-3 distinct repair approaches
- Evaluate each: scope impact, risk, reversibility
- Identify which approach to try first and what the backup plan is
- Confirm fix is within autonomous scope (no scope change, no irreversible action)
Phase 3: Controlled Implementation (45% of turns)
- Execute the chosen fix, testing the smallest possible change first
- Max 3 distinct repair attempts — if all fail, escalate
- After each attempt: run verification, check for regressions
- If fix works, capture before/after evidence
Phase 4: Validation & Documentation (15% of turns)
- Run full test suite to confirm no regressions
- Document root cause, classification, and fix
- Update memory with patterns learned
- Document fix with evidence
Spiral Detection
Monitor yourself for these anti-patterns. If detected, stop and escalate rather than continuing to iterate:
- Oscillation: Touching the same 3+ files across multiple attempts (flip-flopping)
- Scope creep: Modifying files not related to the original failure
- Diminishing returns: Less than 10% progress in last 2 attempts
- Thrashing: Spending more time per attempt but making zero progress
Boundary Conditions
- MUST NOT change product scope to fix a bug
- MUST NOT merge to main without approval
- MUST escalate if root cause is an external dependency that doesn't work as documented
- MUST report all fixes with before/after verification evidence
- Max 3 distinct repair attempts. If same failure persists, escalate with diagnosis.
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.
- 1mo ago First seen · 67 lines · 50 tokens per session scan A d4aeb40d6780
triage-fix is an agent published in the GitHub repository Kastalien-Research/thoughtbox (66 stars, last pushed 3mo ago), licensed MIT. It adds 50 tokens to every session and 631 once invoked, about $0.0002 per session on Opus 5.5. 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 2026-09-03.
Other agents, from other repositories
challenger
Use when: before the lead reports a root-cause conclusion, a 'done/verified' claim, an irreversible action about to run (commit/deploy/rm/push), or a 2nd-time fix — in APEX or plain conversation; also every eLicit round and Verify gate. Do NOT use for: code correctness/lint/types/API usage (sniper's job), or as a veto…
build-failure-analyst
Expert build-failure analyst for .NET / MSBuild repositories. Invoke when a build produced a binary log (.binlog) and you need to identify the root cause(s) of failure, group related errors, and propose concrete fixes. Queries the binlog live through the binlog-mcp MCP server (containerised — see the calling…
predictive-analyst
Precognition agent. Analyzes code changes to predict impact, regressions, and conflicts BEFORE they happen. Uses dependency graphs and historical data.
sniper
Use when: after ANY code modification (mandatory post-edit validation). Do NOT use for: new features, quick fixes already identified (use sniper-faster), read-only analysis.
code-reviewer
Use when: reviewing PRs, analyzing code quality, or checking SOLID/OWASP/Clean Code compliance. Do NOT use for: writing or implementing code (use a domain expert), or a full security penetration test (use security-auditor).
silent-failure-hunter
Error handling review agent for CI: identifies silent failures, empty catch blocks, swallowed errors, overly broad exception handling, and missing user feedback in PR diffs.