blog-research
Skill0 starsChanged 3 months agoMITView on GitHub
Use when the user asks to check what a specific blogger/expert has written, find tutorial or how-to content, or research opinion/practitioner writing on a topic. Triggers on "check their blog", "find a tutorial for", "what do practitioners say", "blog research". Produces evidence records (URL + verbatim quote + author credibility note), never conclusions. Do NOT use for official company documentation (use official-source-research), news coverage (use news-research), or general web pages with no identifiable author (use web-research).
In this repository10 other files it gives its agents.
- academic-researchSkill
- github-researchSkill
- news-researchSkill
- official-source-researchSkill
- reddit-researchSkill
- research-intelligenceSkill
- research-plannerSkill
- social-researchSkill
- source-verifySkill
- web-researchSkill
The file
We split this skill into the rules it gives an agent
1 rule. About 0 are common, 0 are distinctive, and 1 is very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈0
Distinctive · estimate
This author's own choices. Worth a look.
1
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 18
Don't use for: an organization's own official docs/announcements (→ official-source-research — same publisher, different tier by default), …
≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Blog Research
Overview
Collects evidence from individual and organizational blogs: practitioner write-ups, tutorials, opinion pieces, an organization's own technical blog. Distinguishing job: blog content ranges from tier 2 (an org's official blog, on-claim-about-that-org) to tier 5 (anonymous content-farm posts) — author identity resolution is most of the work here.
Layer rule: produces evidence, not results. Synthesis and confidence status are the orchestrator's job.
When to use
Direct use: "what does [named expert] say about X", "find a tutorial for Y". As a component: orchestrator dispatches this for implementation/opinion sub-questions where a practitioner's direct account matters more than official docs.
Don't use for: an organization's own official docs/announcements (→ official-source-research — same publisher, different tier by default), news coverage of an event (→ news-research), pages with no identifiable author or track record (→ web-research, capped at low tier there too).
Workflow
- Resolve author identity first — this decides everything downstream. Look for: a named author with a bio, a track record (cited elsewhere, known conference talks, GitHub presence), or organizational affiliation stated as the post's byline. Log what was found in
limitationsif identity is thin or absent — an unattributed post is not automatically discarded, but it caps at tier 4-5. - Distinguish original from syndicated/aggregated. Many blog platforms mirror or aggregate posts (dev.to cross-posts, Medium republishes, translated mirrors). If the passage attributes the content elsewhere or matches a known original almost verbatim, route to engine/origin-tracing.md before treating this post as the source.
- Search via web-research's chain (WebSearch + WebFetch — this skill doesn't need a separate tool chain, the distinguishing work is the author/originality resolution, not the fetch mechanism). If nothing surfaces, run references/query-relaxation.md.
- Extract publish date per the cascade in references/tooling.md — blog platforms are usually good about structured metadata (dev.to, Medium, Substack all expose
article:published_time). - Archive per references/citation-rules.md — personal blogs and smaller platforms have real disappearance risk (platform shutdowns, domain expiry), archive proactively.
- An individual is one root:
root_id: person:<canonical-handle>. The same author's blog post, conference talk, and newsletter are one voice — quoting three of them is not three independent sources. For an org's official blog, useorg:<name>instead (engine/origin-tracing.md). - Assign tier per references/source-tiers.md's blog table (2 = org's own official channel on a claim about that org; 3 = individual with demonstrated expertise; 4 = anonymous/unverified; 5 = content-farm pattern) and purpose — tutorials/how-tos are usually
implementation, stated positions areopinion, factual claims about a product/event arefactonly when backed by the author's own primary access (they built it, they tested it) rather than repeating something they read elsewhere. - Write the evidence record per engine/evidence-contract.md,
verification_status: pending. Do not self-verify.
Output
One evidence record per claim, appended to the run's evidence.jsonl.
Common mistakes
| Mistake | Fix |
| Treating a syndicated repost as an independent source | Trace to the original per engine/origin-tracing.md before tiering |
Assigning fact purpose to a tutorial author's untested claim | Purpose is fact only when the author has primary access to what they're claiming |
| Skipping author-identity resolution because the content "sounds authoritative" | Tone is not a credibility signal — check for a real byline, track record, affiliation |
| Not archiving because the blog "seems established" | Smaller platforms and personal domains disappear; archive regardless |
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

